Recent policy debate in Europe suggests that a shorter workweek will lead to more jobs (worksharing). We derive and estimate a model where the firm employs two types of workers, some working overtime, the rest standard hours. Worksharing is not always a prediction of the theory. Using German establishment-level panel data (the IAB-Establishment panel), 1993-1999, we find no evidence of pro-worksharing effects except in small plants in the East German non-service sector. There is evidence that a cut in standard hours lowers the proportion of overtime workers in a plant, as predicted by the theory, and increases the proportion of standard-time plants. © Scottish Economic Society 2005.