Various aspects of locomotor activity were measured, using a crossed photobeam cage, in young rhesus monkeys (M. mulatta) whose mothers had been fed low levels of polychlorinated biphenyls (PCBs, Aroclor 1248) both before and during gestation and nursing. In Exp. 1, a group of young monkeys whose mothers had been fed 2.5 ppm PCBs (the 2.5 ppm concurrent-exposure group) was compared with the control group of Exp. 1 at 6 (Exp la) and 12 (Exp lb) mon of age. In Exp. lla, a group of young monkeys born from the same 2.5 ppm mothers after they had been off their 2.5 ppm PCB diets for periods ranging from 0.5 to 1.5 yr (i.e., the 2.5 ppm post-exposure group) was compared with the control group of Exp. II at 12 mon of age. In Exp. IIb, two groups of young monkeys from groups of mothers fed 0.5 and 1.0 ppm PCBs three times per week (concurrent-exposure) were compared at 12 mon of age with the control group of Exp. II. In the 12-mon tests the mean data and variability of the two control groups were very similar, both in their mean locomotor activity levels in early sessions and in activity at about half that level in later sessions, suggesting between-session adaptation over the course of the experiment. In contrast, both the concurrent and the post-exposure 2.5 ppm groups showed between-session patterns which began at control activity levels but which rose to levels at least three times those of the controls by the final sessions. In the tests conducted at 12 mon of age, locomotor activity for 15 min periods within each daily session were within measurement error of being stable for all groups except the 0.5 and 1.0 ppm groups of Exp. llb. The former exhibited increasing activity throughout each session, a pattern which persisted across all 24 sessions run. On the other hand, the 1.0 ppm group showed within-session decrements which became less pronounced as the experiment continued. This latter pattern was the only one shown by any group which was consistent with the concept of "reactivity". Overall, all of the PCB-treated groups were more active than their controls with no quantitative difference related to PCB dosage. Dose-effect differences appeared in the form of qualitatively different activity patterns between the groups, for which no quantitative description appeared obvious.