In this paper, we report findings from a recent survey of international migration from China's Fujian province to the United States. We take advantage of the ethnosurvey approach as used in the Mexican Migration Project. Surveys were done in migrant-sending communities in China as well as in destination communities of New York City. We derive hypotheses from two strands of recent studiesthe international migration literature and the market transition debate. Our results are in general consistent with hypotheses derived from cumulative causation of migration. However, because of the geographical location of China as compared to Mexico, there are some differences between the two countries in terms of particular migration patterns to the United States. As expected, at the community level, migration prevalence ratio (measuring migration networks) increases the propensity of migration for other members in the community. In contrast, having a household member migrated previously does not increase the propensity of migration of other household members until debt for previous migration is paid off. Our research clearly demonstrates the value of bringing the case of China into the comparative study of international migration. With respect to market transition theory, we find that political power continues to be an important factor in the order of social stratification in the coastal Fujian province.