In this paper, a longitudinal data set covering 5°7o of all Danish wage earners over a 9-year period is used to shed light on the observed wage differentials due to gender. A human capital model is used to isolate the effects of changes in experience, schooling and unemployment, together with other factors. The model is specified as a fixed effect model that takes into account unobserved factors such as individual skills, the tendency to have an extra job and other factors. Despite the observation from macro statistics that women have had the highest observed increases in wage rates, the models show that this increase is mainly due to an improvement in their background characteristics and that men still receive a higher return to their characteristics. The main difference between genders appears to be that female workers do not, in general, get any return to their experience. The estimates also show negative effects on the wage rate of previous spells of unemployment. * Peder J. Pedersen, the labour workshop at the Aarhus School of Business and two unknown referees are thanked for valuable comments. However, only the authors are responsible for any remaining errors. The Danish Research Council of the Social Sciences has provided the necessary funding.