Developmental effects of polychlorinated biphenyls (PCBs) on male reproduction in rodents have been studied previously in several laboratories, however, with inconsistent results. The effects on testicular weights and fertility depend on the test congener or mixture, the dosage, the developmental stage during exposure, and the age of the animals when examined. We have previously reported an increase in testis weight and a decrease in sperm in vitro fertilizing ability in 3,3',4,4'-tetrachlorobiphenyl treated young males. The effect of Aroclor 1254 on sperm fertilizing ability after neonatal exposure, however, was not observed until 45 weeks of age. The objective of this study was to examine the effects of gestational and lactational exposure of Aroclor 1242 (0, 10, 25, 50 and 100 mg/kg/2 days) on testicular development and sperm quality in 16 and 45 week old male mice. Sperm concentration and quality were examined using in vitro fertilization and computer-assisted sperm motion analysis. In agreement with previous studies in rats, Aroclor 1242 caused an increase in testis size and sperm production in 16 week old males, although not significant. These effects, however, were not dose-dependent, but occurred only in the 25 mg/kg treatment group. In vitro fertilizing ability was not affected in 16 week old mice. In contrast, 45 week old mice had smaller testis than control mice. Sperm production and in vitro fertilizing ability was also decreased relative to control mice. Again, these effects were observed in the 25 mg/kg treatment group and not at the higher doses. Testicular gene expression profiles in affected mice were examined using a commercial cDNA array. The effects at 16 weeks of age, however, can not be explained by the changes in gene expression. These results suggest that the fertility of the aging mouse is susceptible to PCB exposure during developmental stages whereas post pubertal mice are resistant. The non-linear dose-response of Aroclor 1242 indicates that effects on male fertility at higher doses can not be extrapolated to lower doses.