Appalachian records turned Elsey into Elza

The surname ELZA is, on the best current scholarly authority, an American respelling of the English surname Elsey, which descends from the Old English personal name Ælfsige, “elf-victory” (ælf “elf” + sige “victory”); that is the position of the 2022 second edition of Oxford’s Dictionary of American Family Names (DAFN), reproduced by Geneanet, FamilySearch, 23andMe and Wiktionary (Geneanet ELZA; Geneanet ELSEY). The same dictionary’s 2003 first edition had instead called Elza a short form of the female name Elisabeth “though the language of origin is unidentified,” so the one authoritative source has changed its mind, and the only competing theories (Alsace, a Germanic “noble” root, a German/Dutch Else) have no documentary support for this spelling (FamilyEducation Elza). The record evidence fits the Elsey derivation closely: the surname is essentially a United States phenomenon of 566 people in the 2020 Census (rank 41,459), about 94% non-Hispanic White, concentrated in Maryland (18%), West Virginia (14%) and Kentucky (13%), and it arose in at least two independent, documented Elsey→Elza shifts, one in Randolph County, (West) Virginia from the 1830s and another among Tennessee-born settlers of Laurel County, Kentucky in the 1870s–80s (NameCensus Elza; Forebears Elza; sortedbyname Elza index 14; Ancestry: James M Elza). The oldest colonial family of any spelling is the Elzey family of Somerset County, Maryland, founded by Capt. John Elzey, who came from Virginia in 1661/62, patented “Almodington” on the Manokin River in 1663 and died in 1664; but no primary record links that family to the West Virginia Elzas, and the pedigree chain that user trees supply is explicitly flagged “unproven” (Archives of Maryland Vol. 54, p. xxxi). Two confounds dominate any search: worldwide “surname” counts of Elza (Indonesia 664, Russia 154, Brazil 129, Israel 57) are almost entirely the feminine given name captured in a surname field, and in the 19th–20th-century Upper South “Elza” was a common male given name (2,942 SSA births 1880–2025, 72% boys, led by Kentucky), so most “Elza” hits in West Virginia and Kentucky records are first names, not family members (Forebears Elza; NameAlmanac Elza). No bearer of the exact surname has reached encyclopedia-level notability; the name’s one lasting public footprint is Elza, Tennessee, named for railroad engineer Paul Elza, which gave its name to Elza Gate, the main entrance to the Manhattan Project city of Oak Ridge (NRHP MPDF 64500613).

Oxford’s dictionary moved Elza from Elisabeth to Elsey

Only one scholarly reference work carries an ELZA headword, and its two editions disagree. The 2022 second edition of Patrick Hanks’s Dictionary of American Family Names (DAFN2), reproduced under explicit Oxford copyright on Geneanet, reads in full: “Elza: Altered form of English Elsey” (Geneanet ELZA). FamilySearch and 23andMe print the same sentence, and English Wiktionary gives “Altered form of Elsey” as its surname etymology (FamilySearch Elza; 23andMe Elza; Wiktionary Elza, 2023 mirror). The 2003 first edition, as reproduced on FamilyEducation under its family-name heading, said something different: “From a short form of the personal name Elisabeth (see Elizabeth), though the language of origin is unidentified” (FamilyEducation Elza). That first-edition attribution rests on FamilyEducation’s known practice of reproducing DAFN 2003 text and on the exact match of its sister entries (Elzey “altered spelling of Elsey,” Elzy “Altered spelling of English Elsey,” Ellzey “Perhaps an altered spelling of Elsey,” Elsea “Probably an altered spelling of English Elsey or German Else. This name occurs chiefly in TN, OH, and VA”) with DAFN house style; neither Oxford Reference nor Google Books could be opened to confirm it directly (FamilyEducation Elzey; FamilyEducation Elsea). In DAFN’s vocabulary “altered form” means an American respelling of an existing surname rather than an independent origin, so under the current position the meaning of Elza is simply the meaning of Elsey.

Elsey itself is well documented on both sides of the Atlantic. DAFN2, together with the Oxford Dictionary of Family Names in Britain and Ireland (2016), derives it “from the Middle English personal name Alfsi, Elsi (Old English Ælfsige, from ælf ‘elf’ + sige ‘victory’)” (Geneanet ELSEY). The Dictionary of Medieval Names from European Sources attests the Old English name as “Ælfsi” in 967 and gives the same two elements, and a Bishop Ælfsige of Winchester (died 959) shows it in use among the pre-Conquest elite (DMNES Alfsy; Wikipedia Ælfsige). Harrison’s Surnames of the United Kingdom (1912) adds a second, locative English source: “belonging to Elsey (Lincs),” a hamlet in Mumby parish three miles southeast of Alford, whose second element is Old English ēg “island, riparian land” (Forebears Elsey; PastPlace Elsey). The medieval paper trail runs Alfsi/Alsi in Domesday Book (1086), Reginald Elsi in the Yorkshire Templars’ records (1155–77), Elfsi in the Kent Pipe Rolls (1191), Elsy in the Lincolnshire Pipe Rolls (1230), and Walter Elsey, Lawrence Elsie and Robert Elsay in London parish registers of 1552–1590; those dated forms come from a commercial site, SurnameDB, but they cite standard record classes and agree with the Oxford derivation (SurnameDB Elsey). The z-spellings are a phonetic consequence of English itself: Elsey is pronounced /ˈɛlzi/ with a voiced s, and an American clerk writing what he heard produces Elzey, Elzy, Elzie or, with the unstressed final vowel reduced as it is in Appalachian and Southern speech, Elza /ˈɛlzə/.

The competing roots can each be stated and, for this surname, set aside. The Elisabeth route is real as a mechanism, since DAFN documents German, Flemish and Dutch Elsen, Elser and Elsner as metronymics “from the personal name Else, a short form of Elisabeth,” and Elza is a standard feminine given name in Portuguese, Hungarian, Latvian and Georgian (Geneanet ELSEN; Geneanet ELSE; Behind the Name Elza). Under that theory the root meaning is Hebrew Elisheva (אֱלִישֶׁבַע), “my God is an oath,” with “my God is abundance” and “God of seven” as alternative readings of the same consonants (Behind the Name Elizabeth; Abarim Elizabeth). But no source documents a family named Elza arising from a mother named Elza anywhere, and DAFN dropped the idea in 2022. The Alsace theory fails at the dictionary: the Alsace-derived surnames are Elsass, Elsasser, Elsaesser and Elsesser (German and western-Ashkenazic Jewish), from Old High German ali “other, foreign” + saz “seat” (or, per Bahlow, from the river Ill), with Yiddish Elzes and Dutch Elzas, and none of those entries mentions Elza (Geneanet ELSASSER; Geneanet ELSASS). A German-written “Elza” would in any case be read /ˈɛltsa/, because German z is [ts], which is one more reason the spelling is unlikely to have been imported from a German-speaking source (Wikipedia German orthography). The “Germanic for noble” gloss that Wikipedia’s given-name article and baby-name sites attach to Elza is a conflation with the unrelated Frisian male name Elze (surname Elzinga), which DAFN derives from Germanic adal “noble” (Geneanet Elzinga; Wikipedia Elza (given name)). The commercial crest sites add nothing reliable: HouseofNames has no Elza page at all (“this name has Not been researched”), its Elzey page sells a Scottish “Kintyre/Alexander” story that no dictionary supports, and NameCensus’s two origin blurbs contradict each other (“derived from Elizabeth” versus “an Old German root meaning ‘swan’”) without a single source (HouseofNames Elza; HouseofNames Elzey; NameCensus Elza).

Pronunciation follows the tradition. The American surname is EL-zuh /ˈɛlzə/; the Hungarian given name is /ˈɛlzɒ/ with initial stress; European Portuguese /ˈɛɫ.zɐ/ and Brazilian /ˈɛw.zɐ/; German Elsa /ˈɛlza/ and Else /ˈɛlzə/ already carry the voiced s that an English ear spells with z (Behind the Name Elza; Behind the Name Elsa; Wikipedia Hungarian alphabet). Eric Partridge’s 1959 verdict on the forename, quoted by Forebears, still describes the spelling well: “A rather affected spelling (and pronunciation) of Elsa” (Forebears forename Elza).

Four spelling families, only one of them kin

Compiling “variants of Elza” from search engines mixes four groups that have nothing in common but sound. Only the first group is family. DAFN2’s Elsey cluster lists Elsie (“English variant of Elsey”), Elsy, Elcey, Elza, Elzey (“altered form of Elsey”), Elzy, Elzie (“altered form of Elsey, found mainly among African Americans”), Ellzey (“possibly an altered form of Elsey, or of Alzey, a habitational name from a town in Hesse”), Elsea (“probably an altered form of English Elsey or German Else”) and Nelsey (“possibly a pet form of Elsi with a prosthetic N-“) (Geneanet ELSEY; Geneanet ELLZEY; Geneanet ELZIE). American records add Elzea and Ellzea, which the Elsey Family Association treats as the same Virginia lineage, and Ellsea, under which a Preston County, West Virginia Elsey was indexed in 1840 (FamilySearch catalog 1175770; WikiTree Elsey-150).

Group C deserves one concrete footnote because it is the only Jewish connection in the record: Dutch-Jewish families in Amsterdam used both Elzas and Elsas within one household (Sara Elzas, born 1906, married Salomon Elsas in 1928; both murdered at Auschwitz in 1942), which shows that the Alsace name was a living surname in the Netherlands, but nothing connects it to any American Elza (Joods Monument: Sara Elsas-Elzas). Forebears records non-Latin forms for Elza (Georgian ელზა, Russian Эльза, Arabic ألزا), but these are transliterations of the given name, not independent surnames, and no reference work proposes an Arabic, Turkish or Basque origin (Forebears Elza). Several spellings that circulate in casual lists (“Elsah,” “Elzee,” “Ellsey,” “Elzay,” “Helza”) appear in no surname dictionary consulted.

Fewer than 600 Americans carry it, and the worldwide counts are mostly a girl’s name

The hereditary surname Elza is, for practical purposes, an American name of about 550 living bearers. Read directly from mirrored copies of the official Census Bureau surname files, ELZA was counted 569 times in 2000 (rank 36,956) and 534 times in 2010 (rank 40,856), roughly 0.2 per 100,000 Americans, and a secondary site republishing the April 2026 release of the 2020 file gives 566 (rank 41,459) (Census 2000 file, FiveThirtyEight mirror; Census 2010 file, CFPB mirror; NameCensus Elza; Census Bureau 2020 names). The count has been flat for twenty years while the rank slipped, because the number of distinct American surnames grew. Self-reported race is stable at 94–96% White, with “two or more races” rising from 1.2% to 2.8% and Hispanic at about 2% (23andMe Elza). Every sister spelling except Elze and Elzea is two to three times larger, and the family of spellings splits sharply by race in a way that is itself historical evidence.

Sources: official 2000 and 2010 files as mirrored above; 2020 figures from NameCensus pages for Elzey, Elsey, Ellzey, Elzy and Elze, which reproduce the official 2010 rows exactly but were not verified against the Bureau’s 2020 .xlsx.

Taken together these spellings account for roughly 7,000–7,500 Americans, of whom Elza is 7–8%. The racial pattern is the tell: Elza’s composition matches Elsey and Elsea, not Elzy and Elzie, which supports the derivation from the overwhelmingly White English Elsey rather than from the Black-majority Elzy/Elzie spellings. The most plausible reading, which no source states outright, is that one Chesapeake surname produced separate branches: enslaved and free Black families of Maryland and Virginia who took or were assigned the Elzey name (hence Elzey’s 19% and Elzy’s 59% Black share), and White Appalachian families who carried the Elza spelling west.

Sub-national data are thinner, because the Census Bureau publishes surnames nationally only. Forebears, whose 2014 compilation runs about 40% above the Census count for the United States (746), places 18% of American Elzas in Maryland, 14% in West Virginia and 13% in Kentucky, which is striking given that Maryland holds about 2% of the national population; it also reports a leading Maryland share (14%) for Elzey, the strongest single datum for a shared Chesapeake origin of the two spellings (Forebears Elza; Forebears Elzey). Historically the spelling barely existed: Ancestry’s 1840 census summary finds two Elza families in New York, “about 40% of all the recorded Elza’s in USA,” implying about five households nationally, and Forebears counts 92 American bearers in 1880 (Ancestry Elza; Forebears Elza). In 1940 Elza-surnamed men were laborers (38%), farmers (29%), miners (7%) and farm laborers (7%), a coalfield profile markedly more rural than Elsey’s carpenters and salesmen, and Ancestry’s record inventory holds only 109 immigration records for Elza against about 2,000 for Elsey, consistent with a name that arose at home rather than at a port (Ancestry Elza; Ancestry Elsey).

Worldwide figures need the opposite correction. Forebears estimates about 1,874 bearers of the surname Elza in 45 countries, led by the United States (746), Indonesia (664), Russia (154), Brazil (129) and Israel (57), with Israel showing the highest density (Forebears Elza). The same site counts 247,007 people with Elza as a forename, 170,638 of them in Brazil, which is more than 130 times its surname figure, and Namespedia tallies the string as 96% given-name and 4% surname use (Forebears forename Elza; Namespedia Elza). The non-US “surname” rows are artefacts of this: most Indonesians have no family name, so a mononym or given name lands in the surname field of foreign databases; Russian Эльза, Georgian ელზა and the Israeli records are the feminine given name of Russian and Caucasian families; and Brazil’s 129 “surname” bearers sit beside 170,638 women named Elza, the singer Elza Soares (1930–2022) being the archetype (Wikipedia Indonesian names; Wikipedia Elza Soares). Removing those countries leaves roughly 750–800 people, which is the Census count plus Forebears’ usual overcount and small clusters in Canada (7), England (3) and Belgium (4). FamilySearch’s headline of “337,718 records” for the Elza last name and its Family Tree tally (United States 365, Brazil 35, Switzerland 16) are inflated the same way; the Swiss entry is unexplained by any source (FamilySearch Elza). Behind the Name’s surname database has no Elza entry at all and redirects to the given name (Behind the Name surnames: Elza).

The given-name confound is sharper still inside the United States, and it runs in the opposite gender direction from Europe. Social Security data show 2,942 American babies named Elza from 1880 to 2025, 72% of them boys, peaking in 1921 with 85 births, with the boys’ best rank (#343) in 1880; the state file since 1910 is led by Kentucky (195 births, 3.4 times its expected share), Missouri (75), West Virginia (59), Oklahoma (45), Illinois (41), Ohio (39), Texas (30) and Indiana (28) (NameAlmanac Elza). An independent mirror of the SSA national file confirms that Elza sat in the boys’ top 1,000 in 45 years between 1880 and 1935 and in the girls’ only in 1880, 1881 and 1883, while Elzie appears among boys’ names in 65 years (SSA baby-names mirror). Congressman Elza Jeffords (born 1826 near Ironton, Ohio) and Elza Majors (born 1826 in Edmonson County, Kentucky) are early examples (Wikipedia Elza Jeffords; Whalen family: Elza Majors). The practical consequence is that the same Upper South counties where the surname lives are full of men called William Elza Bennett, Elza Staton, Elza Blair, William Elza Jolliffe and Lemon Elza Tucker; every Find a Grave, Fold3 and county-history hit must be checked to see which field the word sits in (FamilySearch: William Elza Bennett; Find a Grave: William Elza Jolliffe). The male given name is most likely a vernacular spelling of Elzie/Elzy, which was itself a common Appalachian boys’ name; whether it fed back into the surname spelling in the same communities is a reasonable inference that no source makes explicitly.

Somerset County’s Elzeys of 1661 are the oldest colonial line

The earliest documented bearers of any spelling in North America are the Elzeys of Maryland’s lower Eastern Shore, and their record is unusually good because it survives in the published Archives of Maryland. Capt. John Elzey “had been settled in Northampton, Virginia, by 1656,” crossed into Maryland in 1661 or 1662, sat on the first commissions of 1661–62 for laying out lands and keeping the peace in what became Somerset County, was appointed presiding justice of a new joint commission on 4 February 1662/3, remained loyal to Maryland during the 1663 Scarburgh incursion from Virginia, lived on his plantation “Almodington” on the Manokin River, and was buried there in May 1664 (Archives of Maryland Vol. 54, p. xxxi; Archives of Maryland Vol. 54, p. xxix). The county court records of 1665–66 show his widow Sarah remarried to Charles Ballard and administering the estate, and a deposition of 16 August 1665 establishes that John’s brother Peter Elzey was owed cattle from the deathbed instructions (Archives of Maryland Vol. 54, p. 616; Archives of Maryland Vol. 54, p. 628). The Almodington tract, patented in 1663, still carries a mid-18th-century brick house built by his grandson John Elzey II, documented in a 1986 National Register nomination (SAH Archipedia: Almodington). Whether John Elzey was English-born or Virginia-born is not stated in any source retrieved. A commercial site’s table, citing Filby and Meyer’s Passenger and Immigration Lists Index, lists John Elzey (Maryland, 1665), Sarah Elzey (1671) and Arnold Elzey (1671); since John was dead by 1664, those dates are headright or land abstracts rather than arrivals (HouseofNames Elzey).

The family stayed prominent for two centuries. Col. Arnold Elzey appears in Somerset judicial records from 1689, lost a servant to death in 1692, held the servant Matilda, a “free mulatto woman,” under repeated court-ordered extensions between 1707 and 1713, served as Naval Officer of Pocomoke in 1708, and died in April 1733 (Archives of Maryland Vol. 106, p. 7; Archives of Maryland Vol. 406, p. 7; Enduring Connections, Heinegg #48; Archives of Maryland Vol. 663, p. 74). A later Arnold Elzey (ca. 1758–1818), at least second-generation native, served in the Maryland legislature and lived in Somerset, Baltimore County and Washington, D.C. (Archives of Maryland Vol. 426, p. 304). The best-known bearer, Confederate Maj. Gen. Arnold Elzey (1816–1871), was born Arnold Elzey Jones Jr. at “Elmwood” on the Manokin and after age 28 dropped “Jones” and formally adopted Elzey “fulfilling his father’s wish”; this is the only deliberate, as opposed to orthographic, surname change documented for any spelling (Wikipedia Arnold Elzey; Somerset County MDGenWeb query). The same county produced Black Elzey families whose records the Nabb Research Center has abstracted: John McClister manumitted Mary Elzey, 24, and her infant son Thomas Dashiell Elzey on 20 December 1826; George Elzey, 30, a free Black laborer, appears in the 1860 census of Tangier District; and Sarah T. Trent Elzey, daughter of Ernest Elzey, died at Fruitland in 1914 and was buried at Dames Quarter (Enduring Connections, Manumissions #172; Enduring Connections, 1860 census #2334; Enduring Connections, death cert. #2644). Elizabeth Elzey’s 1819 will, dividing property among grandchildren named Waters and listing eleven enslaved people by name, shows how the two Elzey populations were bound together (Enduring Connections, Wills #2321).

Two Virginia lines round out the colonial picture. William Ellzey, a lawyer from the Tidewater, signed the Loudoun Resolves for Independence on 14 June 1774, owned 18 enslaved people in 1784, and died in 1796; his will names sons William and Lewis and daughters Lucy, Margaret, Sarah, Catherine Harrison and Ann Russell (Wikipedia Fleetwood Farm; Kentucky Kindred: Will of William Ellzey). In Frederick County, Samuel Elzea (born 1789, son of John Ellzea) married Frances Self in 1813 and founded a Virginia–Missouri family whose five-volume lineage the Elsey Family Association deposited at the FamilySearch Library; two of his sons were farming in Marion County, Missouri by 1850 (FamilySearch catalog 1175770; An American Family History: Samuel LeFevre). That an association calling itself “Elsey” curates an “Elzea/Ellzea” pedigree is the clearest institutional acknowledgment that these spellings belong to one family.

Two Appalachian families independently respelled Elsey as Elza

The surname Elza, as distinct from Elzey, is a 19th-century Appalachian creation, and the record shows it happening twice. The larger family is that of Thompson Elsey/Elza of Randolph County, (West) Virginia. Compiled Ancestry records give him as Thompson Elsey, born about 1805 in Hardy County, Virginia, died 1891 in Randolph County, while the West Virginia marriage index records “Thompson Elza” marrying Sarah White in Randolph County in 1832; his children’s records name the father as “Thompson Elsey Elza” or “Thompson Elza/Elsey” (Ancestry: Thompson Elsey; sortedbyname Elza index 14). From 1859 the sons appear in the Randolph County marriage books under the new spelling: “Soloman Elza” to Virginia White (1859), William H. Elza to Almira A. White (1860), “Lafaett Elza” to America Sandort (1866), Joseph C. Elza to Martha S. Carr (1873), Zachariah Elza to Margaret J. White Farmer (1874) and Adam Elza (1879) (sortedbyname Elza index 13; sortedbyname Elza index 15; sortedbyname Elza index 8; sortedbyname Elza index 4). Solomon “Saul” Elza (1830–1913), William Howard “Tip” Elza (c. 1838–1890), Lafayette “Lafe” Elza (1845–1920, died Harrisonburg, Virginia), Zachariah Taylor Elza (c. 1846–1930), Adam Albert Elza (1854–1934) and Joseph C. Elza (died 1907) raised their families around Dry Fork, Job, Harman, Whitmer, Wymer and Rich Mountain, and the West Virginia vital index records dozens of Elza births there between 1872 and 1910 (Ancestry: Solomon Elza; Ancestry: Lafayette Elza; Ancestry: Adam Elza; sortedbyname Elza index 1). Maxwell’s 1898 History of Randolph County contains, in the portion searchable, exactly one hit for any spelling, a subscriber “Elza, J S, Rich Mountain” (archive.org: Maxwell, Randolph County).

The crucial point for anyone tracing this line is that the father’s generation was Elsey, not Elzey, and that the Elsey spelling survived in a sister branch. User trees carry Thompson back to an Edward Elsey/Elza (born before 1758 in Maryland, died about 1821 in Randolph County, married Sarah “Sally” Thompson) and then to a Thomas Elsey (1760–1841), “priv., Virginia service,” claimed by one submitter as DAR patriot A210116, and beyond him to Thomas Elsey (1714–1773) and Anne Finch (Ancestry: Edward Elsey Elza; Ancestry: Thomas Elsey 1760–1841). Those same trees label a key link “Nicholas Unproven Elsey,” contain internal date conflicts, and cite no deed, will or tax list; the DAR database that could confirm A210116 refused automated access (Ancestry: Nicholas Elsey; DAR GRS A210116). Edward’s son Nicholas Elsey (1798–1883) settled in Preston County, where the family kept the Elsey spelling into the 20th century at Elseys Ridge, Terra Alta, Amboy and Cranberry, buried its dead in the Elsey Farm Cemetery at Salt Lick, and gave its name to Elsey Run (WikiTree Elsey-150; Wikipedia Elsey Run; sortedbyname Elsey index 1). Hardy County marriages of Edward Elsey (1811), Catharine Elsey (1846) and Dorotha Elsey (1846) confirm an Elsey presence on the South Branch of the Potomac, the standard corridor from Maryland into the Alleghenies (sortedbyname Elsey index 9). So the Randolph County Elzas are Elseys from Hardy County who may descend from a Maryland Elsey; a Maryland → Hardy → Randolph route is plausible, but no primary record found connects Thompson Elsey to the Somerset County Elzeys, and the Elsey/Elzey spelling split in the 18th century is itself a reason for caution.

The second respelling happened in the Cumberland Plateau. James M. Elza was born 20 April 1856 in Washington County, Tennessee, the son of James Elsey and Sarah Hulse, and died 25 February 1923 at East Bernstadt, Laurel County, Kentucky; his son William Rufus Elza (1886–1966) died at Corbin (Ancestry: James M Elza). Laurel County records then carry a continuous stream of Elzas (Mamie 1884–1976, Ben 1888–1975, Chester 1891–1982, Charlie W. 1889–1967) through Whitley, Clay and Jackson counties into the 1990s, and an “Elza Cemetery” stands at Pittsburg, Laurel County (sortedbyname Elza index 3; Find a Grave: Elza Cemetery). A third, Tennessee-rooted line is visible but undocumented upstream: William Elza (born September 1847 in Tennessee, died 1923 in Bell County, Texas, son of Thomas S. Elza and Ellender Ann Tillery) heads a Texas family at Temple, Keller and Anson, and Knox County, Tennessee had its own Elzas by the 1890s, among them James Gass Elza (born 1891) and Willie Ray Elza (1899–1989), buried in Knoxville National Cemetery (Ancestry: William Elza list; sortedbyname Elza index 6; sortedbyname Elza index 15). The Midwest shows the spellings in free variation rather than drift: John Elzey (born about 1822 in Ohio) married Maria Pyle in Clinton County, Ohio in 1842 and farmed in Root Township, Adams County, Indiana, where the 1850 enumerator wrote the whole household as “Elza” and later censuses reverted to Elzey (Wolfe Genealogy: John Elzey & Maria Pyle; sortedbyname Elza index 7). The Kentucky Elzys were in place even earlier, with “Elzy Elzy” enumerated in Jefferson County from 1810 to 1840 and Elzy taxpayers in Bullitt County by 1833 (Kentuckiana Genealogy: Elzy). Because the Elza spelling arose separately in at least two families, modern Elzas are polyphyletic even within the English-Elsey derivation; sharing the surname does not imply sharing a 19th-century ancestor.

The migration geography follows three streams. The Maryland Elzeys largely stayed on the Eastern Shore, with a Baltimore cluster by the 20th century. The Randolph County Elzas spilled into Tucker, Pendleton, Barbour, Harrison, Lewis, Monongalia, Pocahontas, Roane and Taylor counties between 1878 and 1935, then followed the classic West Virginia labor migration to Baltimore after the Second World War (Bales Elza Jr., born 1935 at Whitmer, died 2006 in Baltimore, is typical) and to Pittsburgh, Ohio and the Shenandoah Valley (sortedbyname Elza index 2; AncientFaces Elza). That post-war movement, more than any colonial residue, is the likeliest reason Maryland leads the modern state table. The Laurel County Elzas spread to Knoxville, northern Kentucky, Indiana and Texas. Reconstructed from vital-record indexes rather than county census counts, which could not be pulled, the strongholds were Randolph County in 1880 and 1920 (with Laurel County second by 1920), and the 1840 “New York” datum is best treated as a misindexed given name. Canada’s seven bearers and the scattered Pennsylvania, Illinois and Queens, New York Elzas of 1830–1925 have no documented origin, and no passenger, naturalization or Ellis Island record for an immigrant surnamed Elza was found (Forebears Elza; sortedbyname Elza index 1). Military records are similarly thin for the exact spelling: the National Park Service’s Civil War Soldiers and Sailors database returns no results for “Elza,” so any service by the Randolph County brothers of military age was recorded under Elsey, Elsea or Elzy, or did not occur; 20th-century entries include Clinton C. Elza (Army, 1943–45) and Dale Orval Elza (Navy, 1942–45) (NPS CWSS: Elza; sortedbyname Elza index 3).

A railroad engineer gave the name to Oak Ridge’s gate

No person surnamed Elza has an encyclopedia article. Wikipedia’s “Elza” disambiguation page lists only the feminine given name, the codename of Soviet spy Elizaveta Mukasei and the Tennessee community; the complete player indexes of Pro-Football-Reference and Baseball-Reference contain no Elza, and no Elza legislator, judge or author was found (Wikipedia Elza; Pro-Football-Reference E index; Baseball-Reference E index). The one bearer who left a mark did so through a place. Paul Elza was the construction engineer when the Louisville and Nashville Railroad built a bridge over the Clinch River and an underpass near Dossett, Anderson County, Tennessee; materials for the job were marked “Elza” and stacked at a shed he owned by the tracks, and the flag stop took his name. The 1991 National Register documentation calls Elza, “born of the railroad,” the youngest of the pre-Oak Ridge communities, one with a country store but no post office and “never an established community” (NRHP MPDF 64500613; Wikipedia Elza, Tennessee). Nothing further about Paul Elza’s dates, origin or later life was found.

The hamlet’s name outlived it. Manhattan Project site memos of 1942 located the future reservation “near Elza,” General Groves surveyed the ground from “near the hamlet of Elza” in September 1942, and the Luther Brannon House nearby became his first headquarters (Y-12 history overview, DOE; Wikipedia Luther Brannon House). Elza Gate, named like the other gates for a pre-existing community, was the main public entrance to the closed city from 1 April 1943 to 19 March 1949, staffed around the clock by armed guards within a security force of 4,900 civilian guards, 740 military police and more than 400 civilian police; its opening ceremony drew some 10,000 residents, Donald O’Connor and Patricia Neal, and a ribbon cut with power from the X-10 Graphite Reactor (NPS: Elza Gate; Ewing Gallery: Elza Gate Military Police, 1945). The gatehouse is gone, but Elza Gate Park at 105 Oak Ridge Turnpike carries National Park Service wayside exhibits, and the Department of Energy’s “Elza Gate site,” where Manhattan Engineer District warehouses once stored pitchblende and radium-bearing residues, was remediated in 1991–92 (6,700 cubic yards of waste in 818 truckloads) and certified for unrestricted use in 1993 (NPS: Oak Ridge Wayside, Elza Gate; DOE LM: Oak Ridge Warehouses fact sheet). The result is an irony worth savoring: “Elza” is one of the most photographed words of the atomic age, and it got there through a shed by a railroad.

The variant spellings are far better represented among public figures, and the roster maps neatly onto the geography above. Maryland’s Elzeys produced Gen. Arnold Elzey; Ohio’s produced Paul Elzey (1946–1989), a University of Toledo linebacker who played for the 1968 Cincinnati Bengals (Wikipedia Paul Elzey). The Mississippi–Texas Ellzeys include Congressman Lawrence R. Ellzey (1891–1977, Mississippi, 1932–35), Congressman Jake Ellzey (born 1970, Texas’s 6th district since 2021, a retired Navy fighter pilot), combustion scientist Janet Ellzey of the University of Texas and 1960–61 St. Louis Cardinals center Charley Ellzey (Wikipedia Lawrence R. Ellzey; Wikipedia Jake Ellzey; Wikipedia Janet Ellzey; Wikipedia Charley Ellzey). The Elseys are notably English: George M. Elsey (1918–2015), Truman’s Map Room officer and later president of the American Red Cross; Joseph Ravenscroft Elsey (1834–1857), surgeon-naturalist of Gregory’s 1856 North Australian expedition, remembered in the turtle genus Elseya and in Elsey Station, Elsey National Park and the Northern Territory locality and former electoral division of Elsey; Charles Elsey (1882–1966), British champion trainer of 1956, and his son Bill Elsey (1921–2019), who won the 1967 Oaks and 1973 St Leger; plus footballers Earl Elsey and Karl Elsey (Wikipedia George Elsey; Wikipedia Joseph Ravenscroft Elsey; Wikipedia Charles Elsey; Wikipedia Bill Elsey). The Black-majority spellings have their own figures: Amanda Elzy, Leflore County, Mississippi’s first Black assistant superintendent and a founder of Mississippi Valley State University, for whom Amanda Elzy High School was named in 1959; civil-rights activist Johnetta Elzie; and basketball coach Pat Elzie (Wikipedia Amanda Elzy; Wikipedia Johnetta Elzie; Wikipedia Pat Elzie). Minor places include Elsey, Missouri (a post-office name of 1900–1957), Elsey Abbey in Hagen, Germany, and Elzy near Wick, Scotland (Wikipedia Elsey, Missouri; Wikipedia Elsey Abbey; Wikipedia Elzy (disambiguation)).

Everyone else famous as “Elza” bears it as a given name, and the list is worth having precisely so it can be excluded: Congressman Elza Jeffords (1826–1885, Mississippi), Arkansas State coach Elza Renfro (1902–1935), samba singer Elza Soares, Canadian director Elza Kephart, Latvian actress Elza Radziņa and dancer Elza Leimane, Brazilian Army veteran Elza Medeiros, South African soprano Elza van den Heever, Wild Bunch outlaw “Elzy” Lay, lighthouse builder Elzy Burroughs, and, under Elzie, NASCAR’s Buck and Buddy Baker and cartoonist E. C. Segar (Wikipedia Elza (given name); Wikipedia Elza Renfro; Wikipedia Elzy Lay; Wikipedia Elzie). No business, ship, product, fictional character or family association named for the surname was found; a current vessel named ELZA (IMO 9995571) has no documented connection (MaritimeOptima: ELZA).

Where the record is thin, and what to do next

The single most important open question is whether the West Virginia Elzas descend from the Maryland Elzeys. The pieces that would answer it are known but unexamined: the original 1832 Randolph County marriage register (the “Thompson Elza” reading is from an index and may say Elsey or Elsea), Hardy County court and tax records from 1786 onward, the 1787 Monongalia personal-property tax lists published by Heritage Books, and the Elsey Family Association’s volumes at the FamilySearch Library, whose Virginia Elzea/Elsea pedigrees may well reach the Hardy County Elseys (sortedbyname Elza index 14; Heritage Books 116-084; FamilySearch catalog 1175770). The claimed Revolutionary ancestor Thomas Elsey (DAR A210116) needs to be checked in the DAR Genealogical Research System and the Virginia pension index, because if his line is accepted it anchors the family in Maryland before 1760 and converts the “unproven” Edward Elsey link into something testable. The Papenfuse biographical entry for Arnold Elzey (ca. 1758–1818) continues on a page that could not be retrieved and would list his children, the obvious place to look for an Elzey who moved west (Archives of Maryland Vol. 426, p. 304). Genetics could settle in a year what documents may never settle: no FamilyTreeDNA project exists for Elza, Elzey, Elsey or Ellzey, the Guild of One-Name Studies has no registered study, and the Preston County Elsey profile on WikiTree notes “no known yDNA or mtDNA testers” (WikiTree Elsey-150). A Y-DNA test from a male-line Randolph County Elza compared with a male-line Somerset County Elzey (or a Preston County Elsey, or a Laurel County Elza) would show in one comparison whether these are one patrilineage or several English Elsey families that converged on the same American spelling.

Several sources refused automated access and are worth opening by hand. Find a Grave holds the Randolph County memorials (Frank Leslie Elza 1879–1956, memorial 137076466; James Elza 1873–1950, memorial 36359613; Emma Jane Simmons Elza 1893–1963, memorial 148159854) and the Elza Cemetery page, whose cemetery names, inscriptions and photographs could not be read (Find a Grave: Frank Leslie Elza; Find a Grave: James Elza). WikiTree’s profile Elza-37 (Adam Albert Elza, 1854–1934) and its ELZA surname index are robots-blocked but carry cited sources (WikiTree Elza-37). The West Virginia Vital Research Records at wvculture.org, which underlie the sortedbyname index, offer original images for Randolph, Hardy, Preston, Tucker and Pendleton counties. Fold3 and the NPS database should be searched under Elsey, Elsea and Elzy for Randolph County’s units (10th West Virginia Infantry, 3rd West Virginia Cavalry, Confederate 31st Virginia Infantry and 62nd Mounted Infantry), since “Elza” returns nothing and Fold3’s given-name Elzas (Elza Wiley, Elza Chipman, Elza Cain) are noise. The family-history and Civil War chapters of Maxwell’s and Bosworth’s Randolph County histories, beyond the first portion that was searchable, remain unread, as do Hardesty’s 1883 county histories and Cartmell’s Shenandoah Valley Pioneers for the Frederick County Elzeas (archive.org: Bosworth, Randolph County; HSP catalog: Cartmell).

Smaller gaps are worth naming so they are not mistaken for findings. The origins of the Pennsylvania (Lancaster, Pittsburgh), Illinois, Queens and 1840 New York Elzas are unknown, and the Texas line’s Thomas S. Elza has no documented parent. FamilySearch’s sixteen Swiss tree entries are unexplained. Paul Elza of the L&N has no biography; Chronicling America and the Tennessee newspaper archives are the obvious place to find one. The Dutch Familienamenbank, Hungarian and Brazilian official name statistics, and Beider’s Jewish-surname dictionaries were all unreachable, so the non-American “Elza” rows remain inferred rather than demonstrated to be given names, although the inference is strong. And no Elza family association, reunion or one-name study was found in any non-social-media source; absence here is absence of evidence, not evidence of absence.

Conclusion

The deepest lesson of this research is that ELZA is a name made in America by the sound of English rather than carried from anywhere by immigrants, and that its “meaning,” elf-victory, belongs to a pre-Conquest Anglo-Saxon name that the family had already stopped hearing in Elsey by the time Appalachian clerks wrote it down as Elza. That makes the usual family-crest apparatus of Alsace, Hebrew oaths and Germanic nobility beside the point for the surname, however true those etymologies are for the women named Elza in Brazil, Hungary and Latvia; a bearer who wants a root should go to Ælfsige and the 1086 Domesday scribe who wrote Alfsi, and treat the 2003 Elisabeth gloss as a superseded guess. It also means the name is young: five households in 1840, 92 people in 1880, 566 in 2020, with the spelling fixed only after the Civil War, so the whole surname history fits inside a few well-indexed county record offices.

Going forward, the productive questions are not etymological but genealogical and genetic. The documents already in hand establish two independent Elsey-to-Elza families (Randolph County, West Virginia, and Laurel County, Kentucky) and a third Tennessee–Texas line whose upstream is unknown, so a living Elza should first determine which of these he belongs to, and only then ask about Maryland. The Somerset County Elzeys of 1661 are a magnificent documented lineage, but they are, for now, a neighbor rather than a proven ancestor, and the shortest path from “plausible” to “proven” runs through an original 1832 marriage register, a set of Elsey Family Association volumes in Salt Lake City, and a single Y-DNA comparison that nobody has yet performed.