The behavior of five groups of rats (seven rats per group) made hypothyroid for varying lengths of time and one group of seven normal control rats was assessed under forced alternation fixed-ratio (FR1, FR3, FR5 and FR10), alternating lever cyclic-ratio (ALCR) and progressive-ratio (PR3) schedules of reinforcement. Hypothyroidism was produced by adding methimazole (MMI) to the drinking water of pregnant dams from embryonic day E16 to postnatal day P25. Four groups were given replacement thyroxine (T4) injections beginning at specific time points (P1, P7, P13, and P19). There were no differences in behavioral performance between control and experimental groups under the FR schedule, which indicates that the animals' sensorimotor abilities were intact. Under the forced ALCR schedule, all groups reached criteria similarly. However, under the choice lever ALCR schedule, control animals and those which received T4 replacement from early on (P1, P7, P13 groups) performed well and all had reached criteria by 11 sessions. In contrast, animals which did not receive any T4 replacement or received it late (P19 group) took longer to reach criteria and 5/14 animals had not reached criteria at all by 20 sessions. This deterioration in performance was paralleled by an increase in perseverative behavior as evidenced by an increased frequency of pressing the wrong lever when alternation of lever was required. This suggests that congenital hypothyroidism results in increased perseveration leading to a decrease in learning when a discrimination between correct and incorrect operanda is made available.