The catch forecasts, on which annual TACs for many species are based, are made by assessment working groups for the year ahead; i.e., the forecast for year n is made in year n − 1. In making their forecasts these groups also calculate the expected catch for the current year. Thus a second forecast for year n is made during year n. In this paper the “year ahead” and “current year” forecasts are adjusted for changes in fishing mortality and then compared with the actual catch in order to judge how accurate the forecasts are. It is shown that “current year” forecasts are considerably more accurate than “year ahead” forecasts and some conclusions about the causes and consequences of this are drawn.