Summary This study employs the pseudo-panel approach for estimating returns to education in Thailand, while treating the endogeneity bias common to estimates from data on individuals. Pseudo-panel data are constructed from repeated cross-sections of Thailand’s National Labor Force Surveys of workers born during 1946–67. Estimates show a downward bias of the returns to education in least squares regressions with individual data, a result confirmed with instrumental variable estimation. The overall rate of return is between 14% and 16%. Females have higher returns than males, and workers in urban areas have higher returns than those in rural areas.