A year-long double-blind study of 2,3-dihydroxybenzoic acid (2,3-DHB) given orally at a dose of 25 mg/kg four times per day was undertaken in 15 patients with beta-thalassemia major. 2,3-DHB and placebo (mannitol) were tolerated to an equal degree and there were no signs of drug toxicity at the end of 1 year. Efficacy in terms of retardation of iron accumulation could be documented using serial liver biopsies, serum ferritin determinations, or clinical laboratory assessment. Serum iron values increased, as did the iron binding capacity, in the group receiving 2,3-DHB. The increase in iron binding capacity was due to drug interference with the method of determination. Because of the greater efficacy of slow infusions of desferrioxamine in chelating iron when administered slowly, the clinic has shifted its emphasis toward further evaluation of that compound. Nevertheless, in view of the minimal toxicity of 2,3-DHB, further work appears warranted to define its role in the treatment of iron-overload.