This paper reports the preliminary results from an ongoing effort to improve data on bilateral migration stocks, and set out some working hypotheses on the determinants and socioeconomic implications of South-South migration drawing on a survey of the literature. This paper estimates that 74 million, or nearly half, of the migrants from developing countries reside in other developing countries. In other words, South-South migration is nearly as large as South-North migration. Almost 80 percent of South-South migration is estimated to take place between countries with contiguous borders, and most appears to occur between countries with relatively small differences in income. Estimates of South-South remittances range from 10 to 29 percent of developing countries’ remittance receipts in 2005, depending on the allocation rule chosen to estimate bilateral flows. The impact of South-South migration on the income of migrants and natives is smaller than for SouthNorth migration. However, even small increases in income can have substantial welfare implications for the poor, and cross-migration can improve the match between skills and requirements in the countries involved, thus raising efficiency and welfare. The costs of South-South remittances (where such remittances are permitted) are even higher than those of North-South remittances, because of lack of competition in the remittance market, a lack of financial development in general, and high foreign exchange commissions at both ends of the transaction. These findings suggest that policymakers should pay attention to the complex challenges that developing countries face not only as the countries of origin of migrants, but also as destinations. Designing appropriate policies, however, will require considerable efforts to improve data, and careful analysis of the socioeconomic impact of migration on wages, income distribution, gender, health, and migrants’ rights