The present study analyses the problem of adoption of high yielding varieties of rice during the rainfed aman season in Bangladesh. Drawing on information from a village survey, the analysis confirms some of the well-known results (for an incomplete diffusion process) regarding effects of farm size and tenurial status on the decision to adopt and the extent of adoption. In an attempt to probe further the causal relationships behind adoption the analysis was extended to consider characteristics of farm households. Two variables reflecting consumption requirements and family labour availability relative to land were found to contribute significantly to the explanation of variation in adoption and its extent across farm sizes.