How to Break Through a Brick Wall

Most genealogy brick walls aren't missing records — they're wrong assumptions. The ancestor is in the archive under a spelling you haven't tried, in a parish you haven't considered, or with a first name they only used in one document. The methods that actually break walls are unglamorous: recheck what you think you know, search sideways at siblings and neighbors, try every spelling variant, and go offline to records that were never digitised. This page runs through twelve approaches in the order worth trying them. Keeping the surrounding relationships visible often reveals the gap itself — the printable family tree is useful for that.

Have You Rechecked What You Think You Know?

More brick walls fall to this than to any new record.

Verify every fact with an original source. Not an index, not another person's online tree, not a family story. Go back to the actual document. A surprising proportion of walls exist because a date or place was copied wrongly three generations of researchers ago.

Question the birth year. Ages in censuses, marriage records and death records are routinely wrong by two to five years, sometimes far more. If you're searching a tight range, widen it by a decade in both directions.

Question the name. People used middle names, nicknames, anglicised forms and entirely different names at different points in their lives. A William in one record is a Bill, a Liam or a Guglielmo in another.

Question the place. Boundaries moved, parishes were divided, counties were reorganised, and towns were renamed. A birthplace given as one county may be recorded in another archive entirely.

Question the relationship. "Son" in an old record sometimes meant stepson or son-in-law. "Cousin" covered almost any relative. Household members listed together weren't always related as you'd assume.

Question whether it's the right person at all. Two men of the same name in the same parish in the same decade is common rather than rare, and conflating them is one of the most frequent causes of a wall that can't be passed.

Write down which facts are proven and which are assumed. The wall is almost always sitting behind an assumption you've stopped noticing — which is why keeping sources properly, as in how to cite genealogy sources, does real work here.

Have You Searched Sideways Instead of Backwards?

The single most productive technique when a direct search fails.

Research the siblings. Your ancestor's brother's marriage record may name the parents your ancestor's doesn't. A sister's death certificate might give a birthplace. Whole-family research is standard practice for a reason.

Research the neighbors. Families migrated in groups, and the household two doors along in one census is often the same household two doors along in another county ten years later. Following neighbors frequently locates a family that seems to have vanished.

Research the witnesses. Marriage witnesses, baptism sponsors, will executors and bondsmen were rarely strangers. A recurring name across several family documents is usually a relative.

Research the associates. The FAN principle — Friends, Associates and Neighbors — treats the whole social cluster as the research unit. Business partners, fellow parishioners, and people who appear repeatedly alongside your family all carry information about them.

Research the in-laws. A daughter-in-law's family papers may mention your ancestor when your own line's records don't.

Why it works: an individual can be absent from the record, but a community rarely is. When your ancestor isn't findable, the people around them frequently are — and they carry the detail you need.

Have You Tried Every Variation of the Name?

Assume the spelling is wrong until proved otherwise.

Spelling was not standardised. Clerks wrote phonetically, and the same person appears as Smyth, Smithe and Smith across three documents. Literacy was limited, so people often couldn't correct the clerk.

Try systematically:

• Phonetic variants — Wright, Right, Rite
• Dropped and added letters — Hutchings, Hutchins, Hutchin
• Transcription errors from handwriting — a long s read as f, n as u, rn as m
• Anglicised forms of foreign names — Schmidt to Smith, Kowalski to Kowal
• Wildcard searches, where the index supports them
• Searching without a surname, using first name plus place plus date range
• Searching by first name and spouse's name only

Names change deliberately too. Immigrants adopted new forms, people changed names after remarriage or estrangement, and illegitimate children sometimes took a stepfather's surname later in life.

Check the index, then check the original. Indexes contain transcription errors constantly. If a record should exist and doesn't appear, browse the original register page by page for the relevant period. It's slow and it works.

Have You Gone Offline Yet?

The assumption that everything is online is the second most common cause of a permanent wall.

Most records have never been digitised. Estimates vary but the majority of the world's genealogical material sits in physical archives, unindexed and unsearchable from home.

Where to look:

• County and local record offices, which hold parish registers, rate books, school records, poor law records, quarter sessions and manorial documents
• National archives, for military, tax, court and government records
• Church and diocesan archives, particularly for pre-civil-registration periods
• Local history and family history societies, whose members frequently hold transcriptions and indexes that exist nowhere else
• University and specialist libraries
• Newspaper archives not yet digitised, often on microfilm locally
• Cemetery and burial ground office records, which are far more detailed than the headstones

Ask an archivist. This is genuinely underused. Archivists know their holdings, know what's catalogued badly, and will often point you at a record series you'd never have found in a catalog.

Consider a local researcher. For an archive on another continent, paying someone local for a few hours is frequently cheaper than a trip and more effective than years of remote searching.

What Other Record Types Could Reach Them?

When the standard vital records fail, approach from a different direction.

Occupation records. Apprenticeship indentures, guild membership, trade directories, union records, licensing records and employer archives all name individuals and often give ages, addresses and family connections — the approaches are in how to find an ancestor's occupation.

Land and property. Deeds, tax lists, manorial records, tithe maps and land grants place people precisely and often name relatives in transfers between family members.

Court and legal records. Civil suits, criminal cases, guardianship, bastardy bonds, and probate disputes contain enormous genealogical detail and are heavily under-searched.

Military records, including pension applications, which frequently name wives, children and birthplaces in more detail than any civil record.

Poor law and welfare records, which document exactly the people least likely to appear elsewhere — settlement examinations in particular often state where someone was born and to whom.

Religious records beyond baptisms — pew rents, communicant lists, disciplinary minutes, and membership transfers between congregations, which track migration.

School and institutional records, naming parents and addresses.

Immigration, naturalisation and passenger records, which often give a specific place of origin where nothing else does.

The general landscape of what exists is covered in what are vital records, and if you're relatively new to this, genealogy for beginners covers the foundations.

Can DNA Test the Candidates?

Where documents run out, genetic evidence can distinguish between possibilities.

The typical brick wall shape: you have two or three candidate parents for an ancestor and no document that separates them. DNA can often settle it.

How: find matches who descend from each candidate through documented lines. If your matches cluster consistently with one candidate's descendants and not the others, that's strong evidence. Confirming those matches triangulate on shared segments strengthens it considerably.

Test strategically. An older relative — a parent, aunt or uncle — carries more DNA from the relevant generation than you do, and their match list reaches further back.

Fish in more than one pond. Upload raw data to every service that accepts transfers. The match who solves your wall may have tested at a company you haven't joined.

Where DNA won't help: walls more than about six generations back, where shared DNA falls below detectable levels for most descendants, and endogamous populations where everyone matches everyone.

And a caution: DNA suggests, documents prove. A cluster of matches pointing at a candidate is powerful evidence and still needs a documentary explanation of how the connection works. If the search leads toward living people, how to find living relatives covers the forward-tracing side.

Where Should You Start When Nothing Is Working?

If a wall has stood for months, it's worth running through a structured checklist rather than trying the same searches again.

Work down it in order, and write down the answers. A brick wall that survives every row of that table is genuinely difficult; most don't survive the first four.

Set the problem aside if it's still stuck. Coming back in six months with fresh eyes — and with new records digitised in the interval — resolves a surprising number. Record collections grow every year, and a search that failed in 2023 may succeed now.

Frequently Asked Questions

What is a genealogy brick wall?

A point where research stops because a record can't be found or an ancestor's origins can't be established. Most brick walls turn out to be wrong assumptions — an incorrect birth year, a name variant, or the wrong person entirely — rather than genuinely missing records.

How do I find an ancestor with a very common name?

Research the whole family group rather than the individual, and use neighbors, witnesses and associates to narrow candidates. Occupation, exact address and spouse's name are usually what separate two men of the same name in the same parish.

What if my ancestor's records were destroyed?

Look for substitutes. Tax lists, land records, church records, newspaper reports, court cases, poor law documents and school registers all place people and frequently name relatives. Destroyed record sets are rarely the only evidence a person existed.

Should I search for spelling variants?

Always. Spelling wasn't standardised, clerks wrote phonetically, and indexes contain transcription errors. Try phonetic variants, dropped letters, handwriting misreadings, wildcards, and searching by first name plus place with no surname at all.

Can DNA break a genealogy brick wall?

Often, within about six generations. Where you have candidate ancestors and no document separating them, matches clustering with one candidate's documented descendants is strong evidence — best confirmed by triangulating shared segments.

Why can't I find anything online?

Because most records aren't online. The majority of genealogical material sits in physical archives, unindexed. County record offices, church archives, local history societies and national archives hold enormous amounts that no search engine reaches.

Two habits help more than any single technique. First, write down what you've already searched — which record sets, which spellings, which date ranges — so you stop repeating failed searches without noticing. Second, state the wall as a specific question rather than a vague frustration: not "I'm stuck on Thomas" but "which of the three Thomas Barlows baptised in Lancashire between 1798 and 1812 married Ann Shaw in 1824?" A precise question is often halfway to an answer.

Work through it in order: recheck your assumptions, then search sideways at siblings and neighbors, then exhaust the spelling variants, then go offline. DNA comes last and confirms rather than discovers. Most walls that fall do so because someone questioned a fact everybody had stopped questioning. Keep the surrounding family visible while you work with the printable family tree.