The hypothesis that the banking system consists of firms that use the same production technology is tested and rejected in this study. Six groupings of the population of commercial banks are identified using cluster analysis. The banks are grouped to reflect similar production technologies within groups but different technologies across groups as defined by the strategic conduct (i.e., activities) of the banks. The results suggest that banks in different clusters employ production processes that feature different degrees of substitutability between factors of production, and that the estimates of input substitutability for those groups look quite different from those estimated based on the full population of commercial banks. The impact of the homogeneity production technology assumption on the measurement of cost efficiency is also assessed. The results show that partitioning the industry by strategic conduct reduces the average inefficiency in the industry. These results support those found by others who used similar partitioning criteria but a more narrowly defined sample of banks.