There is evidence that polychlorinated biphenyl (PCB) congeners have differential effects on endpoints of neurotoxicity depending on their chemical structure: specifically, that ortho-substituted congeners are neurotoxic whereas coplanar (dioxin-like) congeners are relatively inactive in producing neurotoxic effects. The effects of the coplanar congener 3,3',4,4',5-pentachlorobiphenyl (PCB 126) on developmental endpoints, hematology, serum biochemistry, and performance on a spatial delayed alternation task were assessed in Long-Evans rats. Dams were dosed with 0, 0.25, or 1.0 microg/kg/day Monday to Friday beginning 5 weeks before and continuing through gestation and lactation. The first 2-week breeding period produced 10, 8, and 13 litters in the three dose groups, respectively. Breeding females from the control and low-dose group that did not conceive were rebred after 76 days of dosing, producing 7 and 6 litters, respectively. Reduction in weight gain from birth to weaning at 21 days of age (DOA) was observed in both dose groups of Cohort 1 but not in Cohort 2. Males in Cohort 1 exhibited a slight decrease in anogenital distance normalized for weight. Changes in hematological and some serum biochemical parameters were observed in the pups at DOA 21 and/or 60. PCB 126 was detected in fat sampled at both DOA 21 and 60. PCB 126 was not detected in brain samples at 60 DOA in any group; analysis of Cohort 2 at DOA 21 revealed levels in the treated group about 1/100 of those in fat. On the spatial delayed alternation task, there was no convincing evidence for impairment as a result of PCB exposure, as assessed by overall accuracy of performance and measures of perseverative and other types of inappropriate responding. These data provide further evidence for the lack of neurotoxicity of dioxin-like PCB congeners. However, assessment of performance on additional behavioral indices is required before definitive conclusions may be drawn.