We evaluate causal impacts of a large-scale agricultural extension program for smallholder women farmers on technology adoption and food security in Uganda through a regression discontinuity design that exploits an arbitrary distance-to-branch threshold for village program eligibility. We find eligible farmers used better basic cultivation methods, achieved improved food security. Given minimal changes in adoption of relatively expensive inputs, we attribute these gains to improved cultivation methods that require low upfront monetary investment. Farmers also modified their shock-coping methods. These results highlight the role of information and training in boosting agricultural productivity among poor farmers and, indirectly, improving food security.