The relationship between health and migration has always been an important theme in immigration research. This research develops a new approach to test the healthy migrant hypothesis and the salmon bias hypothesis in China by examining an interaction term combining agricultural hukou and migrant status, non-agricultural employment history, and subsequent area of residence. Based on two Chinese micro-databases, CGSS 2015 and Harmonized CHARLS, we conducted an empirical test on the relationship between migration and health. Our empirical evidence suggests that the initial health advantage among Chinese rural migrant workers was largely due to self-selection rather than migration effects. After controlling for demographic and socioeconomic characteristics, this advantage disappeared. After their health deteriorated, migrant workers returned to their original location. This could exacerbate the contradiction between the allocation of medical resources and the demand in rural and urban China, further intensifying the already widening health status gap between rural and urban residents.