This paper measures profit efficiency and examines the effect of access to microfinance on the performance of rice firms in Bangladesh. An extended Cobb-Douglas stochastic frontier profit function was used to assess profit efficiency and profit loss of rice farmers in Bangladesh in a survey data of 360 farms throughout the 2008-2009 growing seasons. Model diagnostics reveal that serious selection bias exists that justifies the uses of sample selection model in stochastic frontier models. After effectively correcting for selectivity bias, the mean profit efficiency of the microfinance borrowers and non-borrowers were estimated at 68% and 52% respectively, thereby suggesting that a considerable share of profits were lost due to profit inefficiencies in rice production. The results from the inefficiency effect model show that households’ age, extension visits, off-farm income, region and the farm size are the significant determinants of inefficiency. Some indicative policy recommendations based on these findings have been suggested.