Objective: To assess possible secular trends in sperm variables in men according to previous or subsequent paternity., Design: A comparison of secular trends in sperm variables 1975-1994 in men under investigation for infertility in groups according to previous and subsequent paternity, based on data from the Medical Birth Registry of Norway., Setting: University hospital, Sperm laboratory and the Medical Birth Registry of Norway., Material: Samples were obtained from: 1108 men who had fathered at least one child before the analysis. 1786 men who had at least one child after the analysis, and 2286 men with no children registered., Main Outcome Measures: Sperm concentration, seminal fluid volume, total number of spermatozoa per ejaculate., Results: Analyzed by year of evaluation, men with no child registered had a significant secular decrease of sperm concentration, total number of sperms per ejaculate and seminal volume. The group with subsequent children had significant secular decrease of sperm concentration and total sperm count per ejaculate, while no significant changes were found for the group with previous children. Analyzed by year of birth, a significant decrease of sperm concentration and total sperm count was found for the group without and for the group with subsequent children., Conclusion: The results are compatible with a cohort effect, exhibiting a trend of deteriorating sperm variables in consecutive birth cohorts.