The predictability of corporate profits has attracted considerable research over the last decade or so. While early work focused on the performance of forecast models, corporate managements and financial analystspm se, later studies have increasingly endeavoured to determine the comparative accuracy of the forecasts generated by these sources.1 One reason for this shift in emphasis has obviously been the search for an external benchmark against which to evaluate forecast accuracy. The need for such a benchmark may have become more apparent because of the trend toward profit forecast disclosure by firms in severalcountries, both on a voluntary basis and as a result of policy shifts by accounting rulemaking bodies.' The question then arises whether these corporate forecasts are useful for outsiders -for example, to improve their investment decisions. Financial analysts have served as a knowledgeable group of outsiders in this context, while forecasting models have served as tools potentially available to any outsider.It is, of course, an empirical issue which forecast model could best serve as an external standard for evaluating management's forecast accuracy. This paper tries to shed some more light on this question. It reports primarily on the forecast accuracy of nine models which have been used to describe the time-series of corporate revenues and profits and compares the performance of these models with that of management and analyst^.^ The present research provides several extensions to and replications of previous studies: -whereas earlier research focused almost exclusively on profits, results are given here for both revenues and profits;-the comparisons are based on internal data confidentially obtained from corporate management and financial analysts rather than on published data;