Research Guide

How to Read Old French Handwriting

By François · Racines Genealogy · Published July 24, 2026

It happens to everyone. You followed the guides, you found your ancestor's village, you opened the departmental archives website, you navigated to the right parish register, and there it is: the page that should contain your great-great-great-grandfather's 1780 baptism. And it looks like scribbles. Loops and slashes and brown ink bleeding through from the other side of the page, in a language you may not read, in a handwriting nobody has used for two centuries. Most people close the browser tab at this exact moment.

Please don't. I want to let you in on the best-kept secret of French genealogy: reading old French records is a learnable skill, and a much smaller one than it appears. You do not need to read old French handwriting the way you read a novel. You need to read about thirty words, a handful of fixed phrases, and the names of your own family. The rest of this article shows you how. (New to French research? Start with the complete guide to tracing your French ancestry, then come back here when you meet your first handwritten page.)

The secret: you are reading a form, not prose

Here is the mental shift that changes everything. The priest who wrote that 1780 baptism was not composing. He was filling in a formula he had written hundreds of times before, the same formula his predecessor used, and his predecessor's predecessor. Church rules required registers to be kept in a prescribed manner, and habit did the rest: for centuries, French baptism, marriage and burial entries follow the same skeleton, word for word, with only the names, dates, places and occupations changing.

That means you are not decoding free text. You are reading a form filled in by hand. Learn the skeleton once and you will know, before you decipher a single letter, roughly what each line of the entry is going to say. Your job shrinks to picking out the variable parts: who, when, where, and what they did for a living. That is a task a motivated beginner can genuinely do, and the moment it clicks is one of the great pleasures of this hobby.

The formulas, so you know what's coming

Before 1792, the records you will read are Catholic parish registers. After 1792, when the Revolution created civil registration for everyone, they are civil acts written by the mayor or his deputy. Both are ruthlessly formulaic. Here are the skeletons, with translations.

Baptism (baptême)

Le [date] a été baptisé(e) [name], fils/fille de [father] et de [mother], son épouse. Le parrain a été [godfather], la marraine [godmother]. In English: "On [date] was baptized [name], son/daughter of [father] and of [mother], his wife. The godfather was [godfather], the godmother [godmother]." A fuller real entry might read: Le vingt janvier mil sept cent quatre vingt a été baptisé Jean, fils de Pierre Martin, laboureur de cette paroisse, et de Marie Dubois, son épouse. Le parrain a été Jean Dubois, la marraine Anne Martin. Notice how much of that you can now predict: everything except the five names, one date, and one occupation.

Marriage (mariage)

Le [date], après la publication des bans faite par trois dimanches consécutifs, ont été mariés en cette église [groom], fils de..., et [bride], fille de... "On [date], after the publication of the banns made on three consecutive Sundays, were married in this church [groom], son of..., and [bride], daughter of..." Marriage entries are the genealogist's favorite, because the formula obliges the priest to name both sets of parents, often with their parishes of residence, and to say whether each parent is living or dead. One marriage act can hand you six people at once.

Burial (sépulture)

Le [date] a été inhumé(e) dans le cimetière de cette paroisse le corps de [name], âgé(e) d'environ [age] ans. "On [date] was buried in the cemetery of this parish the body of [name], aged about [age] years." Burial entries are the shortest, and the ages in them are often rough guesses, so treat them gently.

Civil registration (after 1792)

Civil acts are wordier but even more rigid, because the law dictated their content. A birth act (acte de naissance) typically opens: L'an mil huit cent douze, le trois février, par-devant nous, maire et officier de l'état civil de la commune de..., est comparu [father]... "In the year eighteen hundred twelve, on the third of February, before us, mayor and officer of civil status of the commune of..., appeared [father]..." The father then presents the child, states the mother's name, and two witnesses sign. Once you have read three of these, you will skim the boilerplate and jump straight to the names. If terms like commune and département are new to you, our short explainer on how France is organized will make the whole filing system click.

The thirty words that unlock everything

Because the records are formulaic, a surprisingly small vocabulary carries you a very long way. Here is the working glossary I would want taped next to any beginner's screen.

Two of these deserve a spotlight. Feu (literally "late") in front of a name means that person had already died: fille de feu Antoine Roux means the bride's father Antoine was dead by the wedding day, which quietly gives you a bracket for his death date. And ledit / ladite ("the said") is how clerks avoided repeating a name; when you see ladite Marie, look back up the entry for the Marie already mentioned.

The letters that fool everyone

Most of what makes old French handwriting look alien comes down to a short list of letter habits. Learn these and half the fog lifts at once.

The long s. Until the early 1800s, the letter s inside a word was often written as a tall character (ſ) that looks almost exactly like an f. The word that looks like "baptifé" is baptisé. The rule of thumb: if reading it as f makes no sense, try s.

u and v, i and j. Older writing treats these as interchangeable shapes. You will meet febvrier for février, and the classic old spelling Jehan for Jean, which sometimes appears looking like "Iehan." Likewise Estienne is the old spelling of Étienne (Stephen). These are not errors; they are simply how the names were written then, and you should record the modern form while noting the original.

The minim problem. The letters i, u, m and n were each drawn as one, two or three identical downstrokes (minims), with no dots and no distinguishing shapes. A word like commune becomes a picket fence of a dozen strokes that could be divided several ways. The cure is context: the formula tells you a place name or the word commune belongs there, so you count strokes with a candidate word already in mind.

Capitals, doubles, and missing accents. Capital letters in old hands often look nothing like their modern printed forms; a capital S, L or G can be an ornate knot you would never guess. Double s could be written as ſs, which looks like "fs." And accents were used sparsely or not at all, so née may appear as "nee" and décédé as "decede." None of this is your ancestor being sloppy. It is simply a different set of conventions, as consistent in its own way as ours.

Abbreviations: the clerk's shortcuts

Clerks wrote the same words all day, and they abbreviated mercilessly. The ones you will meet constantly: led. and lad. for ledit and ladite; Sr for sieur (a courtesy title, roughly "Mr."); Me for maître (used for notaries and lawyers); St and Ste for saint and sainte in place names. Endings were routinely floated up as superscripts, so Jean-Baptiste becomes Jean Bte, and you will see ditto marks or idem when a word repeats from the line above.

Dates: spelled out, and sometimes revolutionary

Dates in French acts are usually written out in words, not digits: mil sept cent quatre vingt trois is 1783 (in the modern spelling, quatre-vingt-trois; old clerks rarely bothered with the hyphens). This feels slow at first, but number words are among the easiest to learn because there are so few of them, and they repeat on every page.

Then comes the surprise. Between 1793 and 1805 the French Republic used its own calendar: twelve renamed months of thirty days each, with years counted from the Republic ("an II," "an III") instead of from the birth of Christ. A birth might be dated le 15 fructidor an III, which converts to 1 September 1795. You do not need to memorize the system; free online converters (search for "French Republican calendar converter") turn any republican date into a modern one in seconds. Just recognize the month names (vendémiaire, brumaire, frimaire, nivôse, pluviôse, ventôse, germinal, floréal, prairial, messidor, thermidor, fructidor) so you know what you are looking at. The republican calendar was abandoned as of 1 January 1806, so any record from 1793-1805 may use it.

Special cases: Latin, German, and the truly old

Latin. Some Catholic parish registers, especially early ones and those in border regions, are in Latin rather than French. Do not panic: they follow the same formulas, and a tiny Latin vocabulary suffices. Natus (born), baptizatus (baptized), sepultus (buried), conjugati sunt (were married), filius / filia (son / daughter). The names themselves are often Latinized too: Petrus for Pierre, Joannes for Jean, Maria for Marie.

German script in Alsace-Moselle. If your family comes from Alsace or the Moselle, records from the German period (1871-1918), and many Protestant registers before that, are written in German, often in the old Kurrent handwriting, which is a genuinely different alphabet to learn. That is a subject of its own, and our guide to finding your Alsatian ancestors walks through the region's three languages and where each one appears.

The truly old hands. I will be honest with you: as you push back into the 1600s and 1500s, French handwriting gets objectively harder. Spelling was unstandardized, abbreviation was aggressive, and some notarial hands of that era challenge trained paleographers. There is no shame in hitting that wall; it is precisely where professional researchers earn their keep. But the 18th and 19th century records that hold most of your ancestors are absolutely within an amateur's reach.

A method that actually works

Here is the step-by-step approach I recommend to every beginner facing their first register.

1. Start recent, work backward. Begin with your most recent French records, from the 1800s if you can, where the handwriting is closest to modern. As you climb back through the generations, the writing style ages gradually underneath you, and you barely notice yourself learning.

2. Read the whole page, not just your entry. Every entry on the page is by the same scribe, using the same letter shapes. Words you can read easily (month names, "fils de," the priest's own signature, repeated village names) become your key to the words you can't. Build yourself a small letter key: find a word you are sure of, and note how this particular writer forms his capital B, his final t, his double s. Then apply it to your entry.

3. Let the formula predict the next word. You know the skeleton now. After fils de comes the father's name; after the father's name, very often his occupation; after et de, the mother. Reading becomes a game of confirming predictions rather than decoding from nothing, and that is a much easier game.

4. Compare entries by the same hand. Stuck on a name? Search nearby pages for other entries mentioning the same family. The same surname written three times by the same priest, in three slightly different inkings, will usually resolve the doubtful letters.

5. Zoom, transcribe, leave blanks. Use the archive viewer's zoom generously; strokes that blur together at page scale separate beautifully at high magnification. Type out a transcription of what you can see, leaving blanks or bracketed guesses for what you can't, and come back tomorrow. It is remarkable how often a word that defeated you on Tuesday reads itself on Thursday.

Free help is everywhere

You are not alone with this. FamilySearch's free research wiki has pages on French genealogical word lists and handwriting, with guides to reading French records. Geneanet's forums have long-running threads where volunteers cheerfully transcribe snippets that stump researchers; posting a cropped image of a stubborn line often gets an answer from a French reader within days. Several French departmental archives publish their own paleography help pages and reading aids (look for "aide à la lecture" or "paléographie" on the archive site), and since the registers themselves are free to browse, practice costs nothing; our tour of the free online French archives shows you where everything lives.

Technology is starting to help too. FamilySearch has been rolling out full-text search with computer-generated transcriptions for some record collections, and similar tools are appearing elsewhere. Treat these the way you would treat a helpful stranger's guess: a wonderful lead, never a source. The machine misreads old handwriting in exactly the ways beginners do (7bre included), so verify every machine transcription against the image with your own eyes before it goes in your tree.

It compounds

Here is the encouraging truth about paleography: it is a compounding skill. The first record might take you an hour. The tenth takes ten minutes. By the fiftieth, you will glance at a baptism entry and simply see the names, the way you see the fields on a printed form. Every entry you fight through is teaching your eyes, and none of that effort is ever wasted, because the same formulas and the same letter habits are waiting for you in the next parish, and the next. The page that looks like scribbles today is a letter from your family. It was written to be read, and with a little patience, you are exactly the person it was waiting for.

François is a French-born genealogist living in the United States. New to French research? Start with the complete guide to tracing your French ancestry, then how to find the French village your ancestor came from and how to find your French grandparents. Latin or French in an American register? See finding the French village from an American church marriage record. Records in German rather than French? That has an explanation: was your ancestor French or German. A reading problem solved in practice: one real family, traced from New Jersey to two Alsatian villages.