Previous studies of rural-to-urban migration have led to difficulties in distinguishing migration antecedents from migration consequences. Failure to control for key factors such as migration history, age at migration, and duration of migration measures make results difficult to interpret. A prospective study conducted in rural Guatemala allowed for comparisons on a large and diverse set of anthropometric, psychological and socioeconomic variables between eventual urban migrants and rural sedentes as early as age 3 years. Results indicate that eventual migrants may have been absent from the village more often than sedentes even at this young age, suggesting that these children may have been part of families with a history of circular migration. Evidence for selective migration (i.e. migration of individuals who are in some way ‘ better off ’than the originating population) was found in two of the four study villages, where eventual migrants were larger and of better socioeconomic status at age 3 than sedentes. Few differences were found in the other two villages. Differences generally persisted subsequent to migration. Patterns seen are indicative of the economic and social systems characteristic of the individual villages: the two villages in which selective migration was seen offer little to deter individuals from leaving, while the other two villages provide greater economic opportunity for residents. Rural regional differences should be carefully considered in future studies of rural-to-urban migration.