ABSTRACTMany states are concerned hat they are losing a disproportionate amount of human capital through interstate migration. This may be happening not only because migrants tend to possess more human capital than nonmigrants, but also because of human‐capital selectivity in interstate migration. Using Public Use Microdata Sample A from the 1980 United States Census, human capital measures are developed and used to test whether states experience human‐capital migration that is significantly different from human migration. Strong evidence demonstrates that human‐capital migration differs significantly from human migration. For example, Arkansas and Vermont are shown to have a net in‐migration of people, but a net out‐migration of human capital. Conversely, the data for Connecticut and Minnesota indicate a net out‐migration of people and a net in‐migration of human capital. This study suggests that traditional models of interstate migration mask the flow of