Data envelopment analysis (DEA) has been widely used in agricultural production efficiency analysis in recent years. Based on the existing state of DEA technology, this paper estimates technical, allocative, economic, and scale efficiency using field-level survey data from a sample of 199 Boro rice farmers in north-central part of Bangladesh for the year of 2010. The results of the study revealed that on average, the farms technical, allocative, economic, and scale efficiencies were 0.93, 0.82, 0.69, and 0.90 respectively. Their existing technical, allocative, economic and scale inefficiencies were 7%, 18%, 31%, and 10%, respectively. In addition, a second stage Tobit regression showed that the variation was also related to farm-specific attributes such as education, family size, seed type, land tenancy, extension services, irrigation machine type, and sources of energy. Although tremendous development has been achieved in crop production in Bangladesh, the evidence suggests that farmers in Bangladesh fail to exploit the full potential of technology, and that input uses might be reduced through the adaptation and spread of improved agricultural mechanization. Farmers in Bangladesh would also benefit from making sure that a better supply of electricity is delivered to them, and being exposed to new varieties of rice which would be more efficient to grow in their particular conditions.