Attitudes toward risk may affect migration in two ways: First, risk-averse people may be discouraged from migrating due to its inherent risks; second, risk-averse people may use migration as a means to escape from an environment with high income variability. Along the latter line of reasoning, this paper presents direct evidence—something which has never been done before—of a positive relationship between risk aversion and migration. The paper uses experimentaltype measures of risk aversion to explain migration histories of respondents; historical weather data are used to generate a proxy of long-term agricultural income variability which is matched to the times and places respondents have been throughout their lives. The results indicate that highly risk-averse women have a higher probability of migrating away from places with high variability; in the case of males, attitudes toward risk do not affect the incidence of migration.