Three experiments reporting instrumental learning deficits following postnatal iron overload in rodents are described. In Experiment 1, NMRI mice were given different doses of iron succinate (0.0, 3.7 or 37.0 mg Fe2+/kg b.w., p.o.) on different postnatal days (PD). In the PD 10-12 group, there were marked disruptions of radial arm maze learning performance in the 37.0 mg Fe2+/kg dose group, and to a lesser extent in the 3.7 mg Fe2+/kg dose group. In Experiment II, newborn NMRI mice were administered Fe2+ (7.5 mg/kg, b.w.) at either PD 3-5, PD 10-12, or PD 19-21, or vehicle (saline). Marked deficits in radial arm maze performance were obtained in the mice administered iron during PD 1012, and to a much lesser extent PD 3-5. In Experiments III and IV, rats were administered two different dose regimes: (i) Experiment III: 2.5, 7.5, 15.0 or 30 mg Fe2+/kg b.w., p.o., at PD 10-12, or (ii) Experiment IV 2.5, 7.5 or 22.5 mg Fe2+/kg b.w., p.o., at PD 10-12, followed 3 months later by behavioural testing. All four dose groups in Experiment III demonstrated marked deficits in radial arm maze learning and retention performance but only the 30 mg/kg group showed hypoactivity. In Experiment IV, deficits in inhibitory conditioning were obtained in the 7.5 and 15.0, but not the 2.5, mg/kg dose groups. The analysis of enzymes involved in oxidative stress indicated that: (1) Formation of thiobarbiturate acid reactive species (TBARS) concentration was elevated in the substantia nigra by both the 7.5 and 15.0 mg Fe2+/kg doses whereas in the striatum there was a decrease. (2) Superoxide dismutase activity was decreased in a dose-related fashion in the substantia nigra but was elevated in the cerebellum. Iron-overload during the immediate postnatal period incorporating critical synaptogenesis seems detrimental for several aspects of functional and neurobiological development.