Purpose: Groundnut farmers in East Africa have experienced declines in production despite research and extension efforts to increase productivity. This study examined how social network structures related to acquisition of information about new seed varieties and productivity among groundnut farmers in Uganda and Kenya. Design/methodology/approach: Data came from face-to-face interviews with a sample of 491 farmers randomly selected from a larger frame purposefully selected to represent farmers who had worked with researchers and farmers who had not, and to represent both male and female farmers. We used social network analysis to visualize and interpret patterns of farmers' networks with regard to information sources, productivity supports and local group affiliations. Findings: Ugandan farmers primarily used weak ties with researchers and extension agents as sources of information. In contrast, Kenyan farmers used strong ties with close associates. For farmers in both countries weak ties were least associated with productivity. Strong ties, natural factors and farmers' own experience with new varieties were most associated with productivity. The majority of farmers had ties to local groups to strategically pool risks and access available resources. Practical implications: Visualizing farmers' social networks enables policy-makers and change agents to identify relevant social relationships that could be utilized strategically to increase the capacities of poor farming communities. Originality/value: The study demonstrates that important differences in social network structures can exist among farmers in similar geographic regions producing similar crops.