A B S T R A C TGender-disaggregated, household survey data for Uganda are used to examine how gendered roles and responsibilities influence adoption of drought-tolerant (DT) maize, a new technology that can help smallholder farmers in sub-Saharan Africa adapt to drought risk. Multinomial logit (MNL) regression results indicate that, compared to men farmers, women farmers have much lower adoption of DT maize, mainly due to differences in resource access, notably land, agricultural information, and credit. Differentiation of women and men farmers by various characteristics reveals that whether a male farmer was younger or older, or poor or non-poor has no significant influence on DT maize adoption; but important differences among different categories of women farmers are identified. For example, the farmer group found least likely to adopt DT maize is young, poor women household heads. MNL results also show that wives strongly influence adoption of DT maize on plots controlled by their husbands. We discuss the implications of study findings for the development of well-targeted and socially-inclusive adaptation policies.