BACKGROUND: Prior studies showing that drought is (positively) associated with Mexico– U.S. migration looked at periods with historically low rainfall. Though this research informs on the long term potential effects of climate change, it may exaggerate its immediate and medium run effects. OBJECTIVE: We examine the association between rainfall variability, precipitation tends, and U.S. migration from rural Mexico during 2005–2009, a period with above-average precipitation, a more rigorous test of the climate-migration nexus. METHODS: We use multilevel models on microdata and municipal-level characteristics from the 2010 Mexican census and state-level precipitation. RESULTS: In contrast to previous research we find that, in states experiencing relative precipitation declines, these rainfall deficits are associated with lower migration. Yet, we find two instances in which lower rainfall is indeed associated with higher migration in these same states. First, low relative precipitation during the secondary maize growing season implied higher emigration levels. Second, rainfall deficits were associated with U.S.-bound migration out of municipalities with stronger migratory traditions. CONCLUSIONS: We argue that international migration is thus a more common adaptation strategy to climate variability only in times of particularly dire or extended droughts, or out of places with well-established migrant networks. COMMENTS: Our results question the notion that climate change will displace a large number of rural Mexicans: only more extreme climatic variability may displace individuals, though future research should look into whether this kind of migration is temporary. Finally, climate migration projections should also consider the evolution of migrant networks.