We investigate gender differences in agricultural productivity using data collected in 2005 from Nigeria and in 2003 from Uganda. Results indicate that lower productivity is persistent from female-owned plots and female-headed households, accounting for a range of socioeconomic variables, agricultural inputs, and crop choices using multivariate Tobit models. These results are robust to the inclusion of household-level unobservables. However, productivity differences depend on the type of gender indicator used, crop-specific samples, agroecological region, and inclusion of biophysical characteristics. More nuanced gender data collection and analysis in agricultural research spanning diverse regions are encouraged to identify interventions that will increase productivity and program effectiveness for male and female farmers.