Spatial learning and memory was assessed in rats following gestational and lactational exposure to specific ortho-substituted PCBs. Time-mated Sprague-Dawley rats were exposed to PCB 28 (2,4,4'-trichlorobiphenyl), 8 or 32 mg/kg/day, PCB 118 (2,3',4,4',5-pentachlorobiphenyl), 4 or 16 mg/kg/day, PCB 153 (2,2',4,4',5,5'-hexachlorobiphenyl), 16 or 64 mg/kg/day, or corn oil vehicle via gavage on Gestation Days 10-16. Litters were culled to eight on Day 2 and weaned on Day 21. Beginning on Day 90, one male and one female from each litter were tested on a working/reference memory task on an eight-arm maze. For each rat, the same four arms were baited throughout training. Animals were tested Monday-Friday, for seven consecutive weeks. No differences in working or reference memory errors were observed. The same animals were later tested on a T-maze delayed spatial alternation task. On each trial, the reinforcer was placed in the arm opposite that chosen by the rat on the previous trial. Intertrial delays of 15, 25, or 40 sec appeared in counterbalanced order. Rats were tested Monday-Friday for three consecutive weeks. The higher doses of all three congeners resulted in slower acquisition by female rats. Males were not affected. PCB-exposed females were impaired at all delays and were not differentially more impaired at longer delays, suggesting a learning or attentional deficit, rather than a mnemonic deficit. These findings demonstrate that perinatal exposure to ortho-substituted PCBs can result in long-lasting deficits in learning and suggest that the effects of PCBs on learning may be sex specific.