This paper studies the determinants of age-specific employment rates among Swedish males, focusing on the effect of education on employment. We use cohort specific data for the time period 1984-1996 covering male cohorts aged 21-45. It is found that aggregate age-group-specific employment rates increase with the proportion of the cohort with an academic degree. Two states of the labour market are then compared; the high employment period 1984-1990 and the downturn 1991-1996. The effect is stronger in the downturn period as compared with the boom period. However, we do not find any strong evidence in favour of the hypothesis that the effect of higher education on employment is declining with age. A measure of relative education is used to capture crowding out effects. The results indicate a significant effect in the high employment period.