Married women who migrate with their families experience relative earnings losses after migration. In this study, we use data from the 1987 Wave of the Panel Study of Income Dynamics to explicitly examine the relative importance of three sources of those losses: labor force participation, hours of labor supplied, and wages. We estimate earnings models with Heckman's sample selection method for each of four years following migration. The results and subsequent coefficient decomposition methods show that labor force exit and a reduction of labor supplied contribute the largest share to the earnings penalty attached to migration for married women. The participation effect, although reduced in size, is significant for three years following migration. The wages of employed married women who migrate appear to be unaffected in any year following migration.