What do I do when I can't find my ancestor anywhere?

First, define the wall precisely

"I can't find anything on him" is not a research problem; "I cannot identify the parents of a man who appears in County X by a certain year" is. Write the question down in one sentence, then list every record you have already searched for it — with dates and search terms, which is why the research log exists. Most walls shrink at this step, because the written list reveals record groups never actually searched, only assumed to be empty.

Audit your coverage honestly

Run the ancestor through the Record Coverage Checklist on our home page. In our experience the usual culprits are the unglamorous groups: nobody searched the land patents, or the county's probate files, or more than one spelling of the name. The National Archives genealogy portal is a useful audit companion on the federal side — scan its list of record types and ask, for each one, "did I actually search this, for every variant, in every relevant jurisdiction?"

Question every assumption

Brick walls are usually made of assumptions, not missing records. The standard ones to interrogate:

• The spelling. Have you truly searched all variants, or just the family's modern spelling? (How to build a variant list.)
• The place. County boundaries moved; the family may also simply have lived one county over from where the story says. Check the boundary history on the FamilySearch Research Wiki and widen the search ring.
• The date. Reported ages drift across records. Widen your birth-year window by several years in each direction and re-run the key searches.
• The identity. Are you certain the records you have already gathered all belong to one person, and not two same-named men merged into an impossible composite? Un-merging a false composite dissolves many walls instantly.

Research the cluster, not the individual

When the direct trail is cold, the people around your ancestor carry the story forward. Siblings, in-laws, witnesses on deeds, bondsmen on marriages, and recurring census neighbors form a cluster that traveled, married, and litigated together. A brother's probate file may name your ancestor as an heir with a stated residence; a sister's death certificate may name the shared parents your ancestor's records never did. Work the cluster's records with the same rigor you gave the ancestor's own — courthouse records are where clusters show themselves best.

Change repositories, not just search terms

If you have only searched the big online databases, you have not finished searching. State archives, county historical societies, and church record collections hold material that was never digitized. The Library of Congress local history and genealogy guides identify these repositories state by state, along with published county histories and surname files where an earlier researcher may have faced — and sometimes solved — your exact problem. Treat published solutions as leads to verify, never as answers to copy.

Know when to rest the problem

Some walls fall to patience: new collections are digitized and indexed every year, so a search that failed before can succeed today. Log the wall, note the date of your last full attempt, set a reminder to re-run the key searches periodically — and go strengthen another line in the meantime. Progress elsewhere in the tree has a way of circling back: the cluster around a different ancestor is sometimes the same cluster, seen from the other side.