Female monkeys were dosed with 0, 10, 25 or 50 micrograms/kg/day of mercury as methylmercuric chloride. When blood levels reached equilibrium, females were bred to untreated males. A total of 5, 1, 2, and 5 live infants were born in the four dose groups, respectively. Infants were separated from their mothers at birth, and dosed with the same dose their mothers had received. Maternal blood mercury levels averaged 0.33, 0.78, or 1.41 ppm for the three dosed groups respectively. Infant blood mercury levels averaged 0.46, 0.93, or 2.66 ppm at birth, and decreased slowly to steady state levels of 0.20, 0.25, or 0.60 ppm. Behavior was assessed during infancy on a nonspatial discrimination reversal task and fixed interval performance, and when monkeys were juveniles on a series of nonspatial discrimination reversal tasks. During infancy monkeys were tested 7 days per week, 16-21 hr per day in a home-cage environment. As juveniles, they were tested five days per week in a standard operant test environment. For the discrimination reversal tasks, there were no strong indications of differences between treated and control monkeys either as infants or juveniles. Treated monkeys tended to perform transiently better than controls when first introduced to the task both as infants and juveniles. On the Fl, treated monkeys received more reinforcements, and had shorter pauses and lower quarter-life values than control monkeys. Analysis of feeding behavior over the session during infancy revealed marginally longer periods of feeding in methylmercury-treated infants. These results suggest that pre-plus postnatal exposure to methylmercury did not result in gross intellectual impairment in these monkeys, but may have interfered with temporal discrimination.