The catabolism of heme, generating biliverdin, carbon monoxide, and free iron, is mediated by heme oxygenase (HO). One form of this of this enzyme, heme oxygenase-1, is inducible by numerous agents which promote oxidative stress, and is now known to provide important antioxidant protection, as demonstrated in many rodent models of free radical-mediated pathogenesis, and suggested by epidemiology observing favorable health outcomes in individuals carrying high-expression alleles of the HO-1 gene. The antioxidant impact of HO-1 appears to be mediated by bilirubin, generated rapidly from biliverdin by ubiquitously expressed biliverdin reductase. Bilirubin efficiently scavenges a wide range of physiological oxidants by electron donation. In the process, it is often reconverted to biliverdin, but biliverdin reductase quickly regenerates bilirubin, thereby greatly boosting its antioxidant potential. There is also suggestive evidence that bilirubin inhibits the activity or activation of NADPH oxidase. Increased serum bilirubin is associated with reduced risk for atherogenic disease in epidemiological studies, and more limited data show an inverse correlation between serum bilirubin and cancer risk. Gilbert syndrome, a genetic variant characterized by moderate hyperbilirubinemia attributable to reduced hepatic expression of the UDP-glucuronosyltransferase which conjugates bilirubin, has been associated with a greatly reduced risk for ischemic heart disease and hypertension in a recent study. Feasible strategies for boosting serum bilirubin levels may include administration of HO-1 inducers, supplementation with bilirubin or biliverdin, and administration of drugs which decrease the efficiency of hepatic bilirubin conjugation. The well-tolerated uricosuric drug probenecid achieves non-competitive inhibition of hepatic glucuronidation reactions by inhibiting the transport of UDP-glucuronic acid into endoplasmic reticulum; probenecid therapy is included in the differential diagnosis of hyperbilirubinemia, and presumably could be used to induce an ''iatrogenic Gilbert syndrome''. Other drugs, such as rifampin, can raise serum bilirubin through competitive inhibition of hepatocyte bilirubin uptake--although unfortunately rifampin is not as safe as probenecid. Measures which can safely achieve moderate serum elevations of bilirubin may prove to have value in the prevention and/or treatment of a wide range of disorders in which oxidants play a prominent pathogenic role, including many vascular diseases, cancer, and inflammatory syndromes. Phycobilins, algal biliverdin metabolites that are good substrates for biliverdin reductase, may prove to have clinical antioxidant potential comparable to that of bilirubin.