Usage Tips
The Ancestral File is a collection of user-submitted pedigrees submitted to the Church of Latter Day Saints from 1978 until several years ago (new submissions are no longer being accepted). Information on the "same" individual from multiple contributors is merged into the same entry, so it is the granddaddy of current projects like OneWorldTree and wiki projects like this one that seek to combine all the available information on a particular person. The database has over 35 million people in it, and is thus a valuable resource representing the research of thousands of people.
The Ancestral File can be searched at using the link at left. You can narrow results using spouse, year of a particular event, state/country, or parents. You can search only for records with certain parents, as well. Records span virtually all surnames, locations, and time periods, but are for obvious reasons concentrated in the United States.
Data Quality
There are several serious weaknesses to the Ancestral File. First, the records are only as good as the submissions, and submitters were not ever held to a particularly high set of research standards or documentation. Second, the merging process is, for lack of a better word, "dumb." That is to say, the process lacks a human check using common sense. Records will often have multiple spouses with the same name, impossible birth, marriage or death dates, or multiple sets of children born during the same time (the result of two people of the same name being improperly merged). Third, there does not seem to have been an effective corrections process permitting people to be disconnected (removed from incorrect parents or children). If an old theory was rejected and the correct parents given in a new submission, the Ancestral File often just shows both sets of parents as "alternates." This means not only is bad original research included, but work that was published long ago and since proven incorrect lives on in the Ancestral File.
As a result, for any given record, the information may not be trustworthy due to errors at several points along the process. Because more recent (late 19th century and up) records will tend to have fewer submitters with more direct documentation and less need to merge, they are probably more reliable. The further back and the more submissions referencing a record, the greater the possibility for error.
It has been possible to download a gedcom directly from Ancestral File (either online or from CDs) for over 15 years. There are thus a number of pedigrees posted out there that consist of nothing but Ancestral File records. For the reasons stated above, these can't be trusted. Isolated citations to the Ancestral File need to be evaluated just like any other piece of information of unknown reliability.
| 
