Many figures ranging from 6% to 20% have been put forward on the average profitability of non-financial firms. These deviations are due to differences in concept, coverage and sources. For example, the estimates based on corporate data are hard to reconcile, in both level and growth, with the national accounts estimates. Profitability is basically a discount rate whereas it is usually measured as a ratio of a flow of income to a stock of capital. Capital price variations need to be incorporated into the profitability measurement if the measurement and the concept are to be consistent. Only the net trading profit can be estimated in an economically pertinent manner using corporate accounts data. This evaluation first requires an estimate of the lifetimes of capital goods to be able to re-evaluate the stock of capital and calculate depreciation independent of tax considerations. The average lifetime of reproducible fixed assets is estimated at 12 years, which is a finding comparable with other studies based on microdata. This reprocessing provides corporate-data-based profitability estimates that are closer to the national accounts estimates than those usually provided using corporate data. Nevertheless, substantial differences remain, especially as regards the stock of capital. As regards growth, the reprocessing of the corporate data and the inclusion of the asset re-evaluations brings the findings based on the two sources more in line with one another. Over the 1985-2001 period, the proposed estimates show that the earning capacity of non-financial firms was relatively stable at approximately 8% aside from a peak in the late 1980s.