# What to do when you hit a brick wall > Nine moves in order for a genealogy brick wall. Most walls fall to one of the first four, and almost none fall to searching the same index again. By Oleg Khmelnytskyi, Wiserta founder (https://wiserta.com/about#founder). Updated 2026-09-23. Canonical page: https://wiserta.com/genealogy-brick-wall > **The short answer** > Most walls fall to one of four moves: re-read the original documents rather than your notes, check that the last proven generation actually *is* proven, trace every sibling forward, and research the neighbours. Almost none fall to searching the same index again. A brick wall is rarely a missing record. It is usually a record you haven't yet learned exists, a spelling you haven't tried, a jurisdiction you didn't know about, or an assumption two generations lower down that quietly went wrong. Below are nine moves in order. Work through them in sequence and resist skipping to the interesting ones — the first four solve most walls, and they're the least exciting. Before you start, do one thing: **write the wall down in a single sentence.** "I cannot find the parents of James Kelly, born about 1861 in County Mayo." If you can't compress it to one sentence, the problem isn't a wall, it's that you have four unfinished questions running at once, and each needs its own page. ## 1. Re-read what you already have Not your notes. The actual documents. Go back to the images and read them completely — every name on every page, including the ones you skipped the first time because they meant nothing then. Witnesses. Informants. Executors. Bondsmen. Godparents. The neighbours on the census page above and below. When you first collected those records you were looking for one name. Now you're looking for a network, and the network was there all along. This move alone solves a surprising share of walls, and it costs an evening and no money. ## 2. Check that the wall is real Confirm that the last generation you consider proven actually is proven. A large number of brick walls are not walls at all — they're the consequence of an assumption made two generations earlier that was never tested. You attached a family to a tree because the name and county matched, and everything above it is now unfindable because you're looking in the wrong family's records. Ask the uncomfortable question: what is my actual evidence that this man is that man's son? If the answer is "the ages fit and they were in the same county", that's a hypothesis, not a proof, and it may be the real problem. ## 3. Research the whole family, not the ancestor Trace every sibling forward, to the end of their lives. This is the highest-yield move in the entire list and the one most people skip, because it feels like a detour away from the person they care about. Younger siblings die later, in eras with better record-keeping. A sibling's death certificate names the same parents. A sibling who stayed put when your ancestor emigrated leaves a full paper trail in the old country. A sister's marriage record names the same father under a spelling you'd never have guessed. You aren't looking for your ancestor. You're looking for anyone who shares his parents and left a better record than he did. ## 4. Research the neighbours The FAN principle: friends, associates and neighbours. People rarely moved alone. If four families appear together in a Pennsylvania county in 1810 and the same four appear in Ohio in 1830, then a record naming any one of them may place your ancestor — and the family your ancestor married into is very often already inside that cluster. Where to harvest names: marriage witnesses, baptism godparents, will executors and witnesses, deed grantors and adjoining landowners, bondsmen, census households on the same page, naturalisation witnesses, pallbearers named in an obituary. One pattern worth knowing: when a man sells land unusually cheaply to someone with a different surname, that person is very often a son-in-law. Deeds that say "for love and affection" almost always mean family. --- **A structured place to work a wall** **Genealogy Starter Tools** — a research log, a source sheet and an evidence-analysis page, free after one email confirmation. A4 and US Letter, printable and fillable. → [Get the free tools](/free/genealogy-starter-sheets) --- ## 5. Change the spelling Names were written by clerks who heard them spoken, often in an unfamiliar accent, and who were not concerned with consistency. Your ancestor did not have a "correct" spelling. He had whatever the man with the pen wrote that day. Things to try: - Search by **first name and place with no surname at all**, within a tight date range. - Wildcards, if the database supports them. - Say the name aloud and write down every way an English-speaking clerk could plausibly render it. - Check whether the given name was anglicised, translated, or replaced entirely on arrival. - Look for a middle name used as a first name — extremely common, and it hides people completely. ## 6. Change the record type If vital records fail, the answer is usually somewhere less obvious: land and deeds, tax lists, probate and wills, church membership, school enrolment, poorhouse and almshouse registers, coroner's records, newspapers, city directories, fraternal and union records, employer records, voter registrations. Land records in particular survive where vital records don't, and they name heirs, in-laws and neighbours explicitly. A useful benchmark: an ancestor who lived a full adult life in one place should leave twenty or more records across all types. If you have three, you are not stuck — you are early. ## 7. Change the jurisdiction Ask which courthouse held the paper **on the day the event happened**, not which one holds it now. Counties were created, divided and renamed constantly. A birth in 1835 may be recorded in a county that no longer exists, or in the parent county it was carved out of. State lines moved. Parish boundaries moved more often than county ones. Also check the neighbouring county and the state line five miles away — people used the nearest courthouse, not the correct one. And find out whether the courthouse burned. Burned counties are common and they have well-documented workarounds — tax lists, land patents, church records, newspaper abstracts. Find the guide for that specific county before you conclude the record is gone. ## 8. Read the local history County histories, church centenary books, local historical society journals, published cemetery transcriptions, and community cookbooks with a history section at the front. These name the same families you're chasing, often with detail that appears in no official record — where they came from, who they came with, which church they attended, which mill they worked at. They are frequently digitised and free, and they are consistently under-used. ## 9. Write it up as if for a stranger Set out the evidence in order, in full sentences, for someone who has never heard of this family. This sounds like a chore and it is the move that most often produces the answer. The gap that has been invisible for five years usually becomes obvious in the writing, because prose does not let you skip the step you've been unconsciously skipping every time you think about the problem. It also converts your five years of work into something another person can pick up — which is worth doing regardless. ## Two things worth writing down before you start again **The assumptions you've been making without proof.** List them explicitly. "I have been assuming he was the eldest son." "I have been assuming the family was Catholic." Half of these will turn out to be inherited from an online tree you looked at in the first week. **What would change your mind.** Decide, in advance, what evidence would make you abandon your current hypothesis. If you can't name it, you aren't testing a hypothesis — you're collecting support for a conclusion you've already reached, and you'll keep finding it. ## What usually goes wrong **Searching the same index again with more hope.** If a search failed thoroughly once, it will fail identically the tenth time. Change a variable or stop. **Working the interesting ancestor instead of the accessible one.** The way through a wall is almost always a sibling, an in-law or a neighbour — someone you don't care about at all. **Treating DNA as a shortcut.** DNA doesn't find ancestors; it confirms or destroys hypotheses you built on documents. Without a paper tree to test it against, a match list is a thousand names and no information. Build the hypothesis first, in writing, then test it. **Not resolving conflicts on paper.** When two sources disagree, write both down, weigh them by who gave the information and how soon after the event, and record the conclusion **and the reasoning**. If you can't resolve it, say so on the sheet: "born 1861 or 1863, unresolved". An honest uncertainty is worth more than a confident guess, and future you will not remember which one this was. --- **The full protocol, on paper** **The Ancestor File** — a 61-page bookmarked system with the nine-move brick wall protocol, a brick wall worksheet, cluster and FAN research sheets, a conflicting-evidence worksheet and a hypothesis-and-proof sheet. US Letter and A4. → [See The Ancestor File](/shop/the-ancestor-file) --- ## Planning research in an ancestral hometown Before visiting an archive or cemetery, confirm the historical place name, current jurisdiction, opening hours and record access. If you want help connecting one family branch to places and repositories, [review the Heritage Journey preparation service](/heritage-journey). Its scope is agreed before payment; a preparation report is not a travel booking or a guarantee that records will be found. ## Where to go next - [Where to start with family history](/start-family-history-with-what-you-have) - [How to read a US census record](/use-us-census-records-family-history) - [Free Wiserta resources](/free) --- *Wiserta materials are research and wellbeing tools, not legal or professional advice. Records access rules vary by country and by state and change over time.*