This paper extends the Schure and Wagenvoort (1999) study, which considers economies of scale and efficiency in European banking, in a number of directions. Firstly, we introduce what we believe to be important improvements to estimating efficiency. Secondly, we examine more closely the characteristics of the banks on the cost frontier and draw comparisons between the characteristics of these banks and those of the relatively inefficient banks in our sample. In contrast with the results of Schure and Wagenvoort (1999), we find that both X-efficient savings banks and Xefficient commercial banks have cut their average cost, independent of input price movements or changes in the output mix, by about 5 percent each year during the five-year period following the implementation of the Second Banking Directive of the European Union on the first of January 1993. In other words, we observe structural shifts in the cost frontier, possibly due to technological progress. X-inefficient banks became only slightly more inefficient than their efficient counterparts. Therefore, these banks also experienced substantial reductions in average cost. However, these reductions are mainly explained by lower interest expenditures because of lower real interest rates rather than structural changes. Additionally to the results of Schure and Wagenvoort (1999), an examination of the output mix of efficient and inefficient banks reveals that the efficient banks are likely to be more involved in more 'commercial' (or fee-based) activities since their ratio of commission revenue to total operating income is higher. The profitability of savings banks appears strongly determined by their cost efficiency which supports the efficient structure hypothesis. The average profitability of efficient commercial banks seems mainly driven by other factors, such as possibly output prices, rather than cost considerations.