The aim of this article is to illustrate the usefulness of automated linking methods for the study of French-Canadian migrations in Canada and the United States. It details the construction of a geolocated longitudinal sample of nearly 30,000 Catholic men from Quebec, derived from the linkage of the Canadian and American censuses of 1850–1852 and 1880–1881. This sample makes it possible to answer many questions about the composition of the French-Canadian diaspora as well as internal migration flows within Quebec. A link between the sample, which is made up of migrants and non-migrants, and aggregate agricultural census data suggests a negative association between the agricultural prosperity of the origin community and the propensity to migrate.