Guides

Civil registration in Europe: where the record starts and who may read it

Two dates decide a European search: the year the state took over registration from the church, and the number of years a register stays closed. Both differ by country, and often by province.

Updated 4 September 2026

Why the start date is the first thing to establish

Civil registration in Europe is largely a Napoleonic export. Revolutionary France secularised the registers in 1792, taking the recording of births, marriages and deaths away from parishes and giving it to the mayor of every commune. Wherever French administration went, the system went with it, and wherever it did not, the change came decades later — or, in a few places, more than a century later.

This matters more than any other single fact, because it tells you which kind of document you should be looking for in a given year. Ask for a civil birth record in a year before the state kept one and you will be told, correctly, that it does not exist, when a baptism register covering the same family is sitting in a diocesan archive. Ask for a baptism after the church stopped being the recorder and you may find a thinner, less reliable entry than the civil one next to it.

The second consequence is that borders matter twice: once for where the event happened, and once for which administration was in charge that year. A family in the same village can appear in French-style registers, then Prussian ones, then a third system, without ever moving.

Rule of thumb: find the start year for the exact place, not the modern country. Provinces annexed early kept registers decades before the rest of the state they now belong to.

Roughly when each system begins

France begins in 1792, and the départements have digitised their registers extremely thoroughly. Belgium follows under French rule from the mid-1790s; the Netherlands and Luxembourg from 1811, with parts of the southern Netherlands earlier. The German territories on the left bank of the Rhine were registering in the French manner from the end of the 1790s, but Prussia only introduced the Standesamt in 1874, and it became compulsory across the German Empire on 1 January 1876.

Southern Italy kept Napoleonic registers from around 1809, while the rest of the peninsula was brought into a single system with unification, from 1866. Spain established its civil register in 1871. Portugal made civil registration general in 1911, having required it earlier for those outside the Catholic church. In the Habsburg lands, church registers kept the civil function far longer: Hungary introduced civil registration in 1895, while in Austria proper the parish books served as the legal record until the late 1930s.

Poland has to be answered partition by partition — Napoleonic-style registers in the centre from 1808, kept in Polish and later in Russian; the Prussian Standesamt in the west from 1874; and in Galicia, church metrical books doing the state's work. Greece brought in compulsory registration in the 1920s, on top of older municipal rolls and male registers. The Nordic countries are the happy exception in the other direction: Lutheran parish bookkeeping was so systematic, and began so early, that it functions as civil registration from the seventeenth century.

What a civil act actually contains

A Napoleonic-style birth act is a small narrative, not a form. It records the date and hour of the declaration, the name and age and occupation and residence of the declarant — usually but not always the father — the date and hour of the birth, the child's names, the mother's full name including her maiden name, and two witnesses with their ages and occupations. Marriage acts are richer still: they normally name all four parents, state whether each is living or dead, and note the previous marriages of a widow or widower.

The feature that turns these registers into a genealogical machine is the marginal annotation. In the French system and its descendants, later events are written back into the margin of the birth act: the marriage, sometimes a divorce, and the death. One birth record can therefore hand you an entire life without any further searching, and it is the single most efficient document in European genealogy.

The finding aid matters as much as the record. French and Belgian registers come with decennial tables — ten-year alphabetical indexes of births, marriages and deaths per commune — and reading the table first, rather than paging through years of registers, is the difference between an afternoon and a month. Dutch and German registers have their own annual and multi-year indexes.

Before the state: parish registers and where they live

Behind every civil register is a longer church one. The Catholic requirement to keep baptism and marriage books dates from the Council of Trent in the 1560s, and in practice good series begin in the seventeenth century, sometimes earlier. Protestant and Orthodox churches kept comparable books. These registers are usually thinner in detail than civil acts — a baptism may give only the child, the parents and the godparents — but they extend the line by two or three centuries.

The critical practical difference is custody. Civil registers eventually go to the state: to a municipal archive, a provincial or departmental archive, or a national one. Church registers frequently do not. In Spain and much of Italy, parish books are still held by the diocese, on diocesan terms, and no state archive can produce them for you. In parts of central Europe large collections have been digitised by diocesan archives themselves and put online, which is why a Polish or Austrian search can suddenly become easy while a Spanish one stays stubbornly physical.

Where the church and the state overlap in time, read both. They were written by different people from different testimony, and the disagreements — an age, a mother's surname, a place of origin — are frequently the clue that separates two families who share a name.

Closure periods: the rule that decides what you can see

Recent registers are closed, and every country draws the line differently. Germany reformed this clearly in 2009: birth registers become public after 110 years, marriages after 80 and deaths after 30, at which point the volumes are transferred to archives and can be consulted like any other record. The Netherlands and Belgium use similar bands, in the region of a century for births and shorter periods for marriages and deaths. France opens most civil status records after 75 years, with deaths effectively open at once. Poland transfers registers to state archives after a hundred years for births and a shorter period for marriages and deaths.

Spain is the outlier that surprises people: the civil register is a working legal register rather than an archive, access is restricted in favour of the person concerned and their relatives, and the reform that governs it moved the whole system towards tighter publicity rather than looser. Italy applies general archival rules on personal data on top of civil registration, so a commune may hold a hundred-year-old act quite happily while a more recent one requires you to demonstrate an interest.

The practical consequence is the same everywhere: the further you go back, the more open the record. Do not start with the generation you know least about and hit the closure wall. Start with the earliest fully open generation, establish the chain, then use certified copies as a direct relative to bridge the closed years — which every one of these countries will issue to a descendant who asks properly.

Data protection law protects the living. Closure periods, not privacy law, are what stand between you and a nineteenth-century register, and they are administrative rules with documented exceptions.

Where the scans already are

France is the best-served country in Europe for free access: nearly every departmental archive publishes its digitised civil registers and parish books on its own site, with the decennial tables, at no charge. The Netherlands aggregates enormous quantities of indexed civil registration nationally, and the Italian state archives publish digitised civil status registers through a single national portal that grows steadily.

Cross-border aggregators cover the rest unevenly. The large genealogy platforms hold microfilm images and indexes for many countries, including record sets that no national portal has; diocesan collections in Austria, Germany, Poland and the Czech lands are published by a dedicated parish-register platform; and the archives dealing with Nazi-era persecution and displacement hold, and have opened, one of the most important personal-data collections in twentieth-century Europe.

What is not online is as predictable as what is. Indexes come first, then the most-requested series, then the rest — so notarial acts, military conscription lists, population registers and emigration papers, which are exactly the records that break a stuck line, usually still require a request or a visit. Search the portals first, but plan for the archive.

The mistakes that cost the most time

The first is searching where the family lived instead of where the event happened. A civil act belongs to the commune where the birth, marriage or death occurred, and hospital births in a nearby town, marriages in the bride's parish and deaths in an institution all move the record somewhere the family never lived.

The second is the border. Whole regions changed state in 1871, 1918, 1945 or 1990, and their registers did not always follow. Registers from the German territories that became Polish or Russian after 1945 are split between successor archives and a central registry office in Berlin that holds duplicates for many of those places — a fact that unblocks a great many stuck German searches and is not widely known.

The third is the name. Registrars wrote what they heard, in the administrative language of the moment, so the same person appears as Jan, Johann and Jean, and surnames acquire and lose prefixes and diacritics between neighbouring entries. Search on the earliest known spelling and on the sound, not the spelling you inherited, and treat a matching set of parents and a matching place as stronger evidence than a matching surname.

Questions people actually ask

Can I order a certificate from another country as a foreign descendant?

In most of Europe, yes, for records that are still with the registrar, if you can show the relationship — typically by supplying your own chain of birth and marriage certificates. Requirements vary from a plain online form to a documented application, and some countries charge. For records already transferred to an archive there is usually no relationship test at all.

The parish register I need is not with the state archive. Who has it?

Very likely the diocese, or the parish itself. This is the norm in Spain and common in Italy, and it means writing to a diocesan archive on its own terms, which are typically an appointment, a modest fee and a specific request naming a person, a parish and a range of years. Vague requests are declined everywhere.

Why do the same ancestor's records give different ages?

Because age was declared, not verified. In birth and marriage acts the age of a parent or witness is what someone said on the day, and it drifts by several years across a life. Trust a recorded date of birth far more than a recorded age, and when only ages exist, treat the record in which the person is youngest as the best estimate, because an age declared early in life has had less time to drift.

How far back does European civil registration realistically reach?

The civil registers themselves stop at the start dates above. The line continues in parish registers, which give you the seventeenth century in most of Catholic and Lutheran Europe and occasionally the sixteenth. Before that you leave vital records entirely and depend on notarial deeds, tax rolls and land records, where families are named but relationships must be argued rather than read.

Somewhere to put what you find

KORA keeps the people, dates, photographs and notes together, searches 20+ genealogy archives from a person's card, and works offline with no account.