The purpose of this research was to estimate the economic, allocative, technical, pure technical and scale efficiencies of rural cardamom farms in the hilly region of Nepal. Data Envelopment Analysis (DEA) was used to estimate the efficiency scores. Additionally, Tobit regression was also used to explain the variation in the efficiency scores related to farm-specific factors. The efficiency scores of 100 surveyed cardamom farms revealed that the economic, allocative, technical, pure technical and scale efficiencies were an average of 52, 58, 90, 95 and 96 percent, respectively. The results indicate there can be substantial reduction in inputs and the same level of output can be achieved with the existing technology. In addition, the Tobit regression analysis showed a positive significant effect on efficiencies attributed to farm-specific variables, such as cropping year, credit use, women participation, training and off-farm income. To enhance the efficiencies of rural cardamom farms in the hilly region of Nepal, major policy implications should include adequate government arrangements of credit, training programs, encouragement of women participants, and increasing off-farm income sources. JEL Classification: Q12