Whether patient age per se is a prognostic factor of significance in prostate cancer is controversial. To investigate this issue age-specific relative survival was analyzed, and the number of years lost due to this disease was calculated in a large and unselected cohort of 6,890 prostate cancer patients diagnosed between 1971 and 1987 in the northern region of Sweden. The tumor grade was derived from filed notification forms, which showed 26.4% well (grade 1), 40.0% moderately (grade 2) and 17.7% poorly (grade 3) differentiated tumors. There was an overrepresentation of grade 3 tumors among the youngest patients. The age-specific relative survival rate did not differ significantly among different age groups and slight differences almost vanished when adjusting for tumor grade. This finding does not support the view that tumors appearing in younger patients are more aggressive per se. However, loss of life expectancy differed significantly among all age classes and in all 3 grades. In patients with grade 1 tumors the years lost due to prostate cancer ranged from 11.0 to 1.2 in the youngest and oldest age strata, even though the relative survival was approximately 0.70 in all age classes. It was concluded that even if relative survival is constant with patient age, the absolute impact of prostate cancer at different ages varied substantially as indicated by loss of life expectancy. This finding might indicate that younger prostate cancer patients should be given more aggressive treatment than older patients.