The effect of selectivity bias on earnings differences due to labor force participation decisions is investigated, using Current Population Survey and Census data. The magnitudes of observed earnings differences by both race and gender are clearly influenced by the presence of selectivity bias.