The right archive depends on the year, not on the map.

Three questions that decide the archive

Which year? Church registers run to the 1940s; Soviet civil registration takes over from 1939–44 in the west and from 1919 in the rest of Ukraine. Records younger than 75 years are still with the civil registry, which serves relatives only and asks for proof of kinship; everything older has passed to the state archive of the region, where archive rules apply and that conversation does not happen.

Which rite? In Galicia a Greek Catholic and a Roman Catholic parish served the same village, often from different places, with a Jewish community in the nearest town. Latin was the language of both Catholic rites, so the rite is read not from the language of an entry but from the names in it.

Which eparchy? Horodenka is in the Ivano-Frankivsk region today, but what matters is that it belonged to the Stanislaviv Greek Catholic eparchy: its bishop's copies of the registers are in Ivano-Frankivsk, not in Lviv, and the Roman Catholic books of the same district are in Warsaw. The mistake is invisible to the person making it, and it costs a year of letters to the wrong country.

Lviv: the Central State Historical Archive (TsDIAL)

The archive most people are sent to, and rightly, for the Austrian period. It holds the bishop's copies of the Greek Catholic registers of the Lviv archeparchy and a good part of the Przemyśl eparchy, Roman Catholic and other registers, and the two great land surveys — the Josephine cadastre of 1785–88 and the Franciscan of 1819–20 — which list every household by house number. Much of the register collection was filmed and can be read on FamilySearch for nothing.

Two limits are worth knowing in advance. Austria made registers compulsory in 1784, and very little is older; anybody selling you a search for seventeenth-century baptisms in a Galician village is selling something that does not exist. What goes deeper is the cadastres, which group people by house rather than by surname — and for that reason separate the five families of one name that a register cannot tell apart. And for parish registers TsDIAL is, for practical purposes, an archive of the Austrian period: most of the interwar books, and everything after them, are in the regional archives.

Lviv region (DALO), and the oblast that disappeared

The State Archive of Lviv Oblast holds the civil registration books for the region, the interwar Polish administration, and the Soviet records from 1939. Part of its registry collection can be consulted electronically.

The trap here is a vanished oblast. From 1939 to 1959 Stryi, Sambir, Drohobych, Boryslav and Truskavets were not in Lviv oblast but in Drohobych oblast, which was then abolished. Their records of those twenty years are arranged under Drohobych, and a search “in the Lviv region” passes straight over them. Your own document usually settles it: a birth certificate of 1941 that says “Drohobych oblast” is telling you where to look. Read the administrative unit off the paper, and date it by the paper's year. Villages changed names as well — Jazów Stary became Staryi Yar — and an index answers only to the form it was given.

Ternopil (DATO)

The archive we work in most. It holds parish registers of both rites for the Ternopil region, and — the thing almost nobody expects — the parish books of the 1920s to 1944 shelved in the fonds of the Soviet civil registry. They are not Soviet forms; they are the same printed Liber Natorum and Liber Mortuorum, kept and signed by the priest, in Latin to 1939 and in Ukrainian after. When somebody tells you the church books of a village “end in 1938”, the continuation is usually here, under a name that gives no hint of it.

DATO has put many of its inventories online, and the archivists' notes in them are worth reading closely: beside a register there is often a list of the villages whose entries it actually contains. If your village is not in that list, an earlier answer of “the surname does not occur in this book” was an answer about the book, not about your family.

The household books of the village councils, from 1940 on, list every farmstead with everybody living in it, year by year. The house number written in a parish register — in the interwar books as hamlet and number — leads straight into them. They are kept apart from the main archive, in a repository that opens two days a week, and the later ones in district offices that answer only written requests. It is not work that can be done by post.

Ivano-Frankivsk (DAIFO), and further out

DAIFO holds the consistory copies of the Stanislaviv Greek Catholic eparchy from 1788 to 1944, the files of eighteen district courts — guardianship cases among them, which prove a father's death when no death record exists — and sheets of the 1921 census, surviving unevenly. A marriage entry for an orphaned bride often cites the court's permission with its place, date and number; that line is a reference to a file, and the file holds the proof of her parents' deaths.

Beyond Galicia, in brief. Volhynia is Lutsk and Rivne, and Rivne has quietly published some registers online that are on no genealogy site. Bukovyna is Chernivtsi. Transcarpathia is a separate chain altogether — the archive is in Berehove, the records are Hungarian and Church Slavonic, state civil registration runs from October 1895, and more of it is free online than for Galicia. Podolia is Khmelnytskyi, and a fire in 2003 at its repository in Kamianets-Podilskyi destroyed a great deal; a union catalogue of surviving registers, published region by region, says what is left, and is the fastest way to prove that a book does not exist.

How to ask, so that the archive can answer

Ask about the book, not the person. To “do you hold the death record of N” an archive answers “no” even when the book is on its shelf, because nobody has looked inside it. To “do you hold the death registers of parish X for 1914–1923, and in what condition” it must answer about its holdings. Same envelope, different reply.

Write in Ukrainian, give the village under both its historical and its present name, say that you will pay the published tariff, and ask for a reply by email. Our free letter generator produces exactly that letter, correctly addressed. Archives here are working through the war; some answer in weeks and some in months, and a reading-room visit is interrupted by air-raid alerts and resumes after them.

A letter reaches what the staff can find from their desk in an hour. It does not reach an unindexed register that has to be read page by page, the second copy in another fond, the court file that has to be ordered and waited for, or the repository that opens on Tuesdays. That is the part we do.

What we do, and what it costs

The assessment is free: the surname, the village, the years. A records check is $140 — which books for your village and years survive, in both rites, in which archive and which country, what is online and what is lost. It is a written answer, and if the books did not survive, you have paid $140 to stop searching, which is the cheapest good outcome a case can have. A day in the reading room is $290: the books ordered and read, the entries photographed, transcribed and translated, with a written account of everything that was checked whether or not your family was in it. Archive fees for copying are the archive's, not ours, and you hear the sum before anything is paid.

What people ask about the archives