Heme oxygenase activity was measured in tissues of rats killed after administration of NiCl2 or Ni3S2. Induction of renal heme oxygenase activity occurred 6 hr after NiCl2 injection (0.25 mmol/kg, sc), reached a maximum of five to six times the baseline activity at 17 hr, and remained significantly increased at 72 hr. Heme oxygenase activities were also increased in liver, lung, and brain at 17 hr after the NiCl2 injection; heme oxygenase activities in spleen and intestinal mucosa were unchanged. The effects of NiCl2 on heme oxygenase activities in kidney and liver were dose-related from 0.06 to 0.75 mmol/kg, sc. Three Ni chelators were administered (1 mmol/kg, im) prior to injection of NiCl2 (0.25 mmol/kg, sc); d-penicillamine partially prevented Ni induction of renal heme oxygenase activity; triethylenetetramine had no effect; sodium diethyldithiocarbamate enhanced the Ni induction of renal heme oxygenase activity (three times greater than NiCl2 alone). Intrarenal injection of Ni3S2 (10 mg/rat) caused induction of renal heme oxygenase activity at 1 week but not at 2, 3, or 4 weeks; no correlation was observed between induction of renal heme oxygenase activity and erythropoietin-mediated erythrocytosis. Hypoxia (10% O2, 12 hr/day, 7 days) did not affect renal heme oxygenase activity. Induction of renal heme oxygenase activity was observed in mice, hamsters, and guinea pigs killed 17 hr after injection of NiCl2 (0.25 mmol/kg, sc). These studies established (a) the time course, dose-effect, organ selectivity, and species susceptibility relationships for Ni induction of microsomal heme oxygenase activity, (b) the effects of Ni chelators, and (c) the lack of relationship between induction of renal heme oxygenase activity and the erythrocytosis that develops after intrarenal injection of Ni3S2.