Ultraviolet radiation (UVR)-induced photoproducts can be measured by a number of methods. The newly developed 32P-postlabelling method is feasible in molecular epidemiological studies due to its sensitivity, specificity and little amount DNA needed. We applied the 32P-postlabelling method to investigate the induction and repair of photoproducts (cyclobutane pyrimidine dimers (CPDs) and 6-4 photoproducts) after UVR in human skin in situ and studied the effects of age, skin type and gender. The study included 30 subjects aged 32-78 years. The photoproduct induction levels varied 7- to 15-fold between the individuals tested. All four types of photoproducts were induced at a higher frequency in the older population (>/=50 years) than in the younger population (<50 years). Individuals with skin type I and II had a higher CPD induction frequency than individuals with skin type III and IV. In both cases, the differences in thymidylyl (3'-5') thymidylyl (3'-5')-2'-deoxycytidine induction reached statistical significant levels (p<0.05). Photoproduct repair rates 24 h and 48 h after UV irradiation showed a large inter-individual variation. No clear effects of age, skin type or gender on DNA repair could be detected. Our data suggest that UV-induced DNA photoproduct levels increase with age.