Eleven patients with beta thalassemia major were entered into the trial of the oral chelator 1,2-dimethyl-3-hydroxypyrid-4-one (L1). Their ages ranged from 17 to 26 years (mean +/- SD, 22.3 +/- 2.7). Six were male and five were female. L1 was administered at an initial daily dose of 42.5 to 60 mg/kg as a single dose. After 4 weeks, the dose was increased to 85 to 119 (102 +/- 10.7) mg/kg for 191 to 352 days divided into either two or four doses daily, except for one patient who developed agranulocytosis after 11 weeks and was taken off the trial. Initial serum ferritin values in the remaining 10 patients ranged between 1,000 and 9,580 (5,549 +/- 3,333) micrograms/L and at end of the trial their mean serum ferritin was significantly lower (4,126 +/- 2,278; P less than .05 using the paired t-test). Urinary iron excretion at a daily dose of 85 to 119 mg/kg administered as two divided doses ranged between 0.14 and 0.82 (0.44 +/- 0.26) mg/kg/24 h. In three patients, the four doses per day schedule caused substantially more iron excretion than the same total dose divided into two. During the course of the trial, several possible adverse effects have been encountered. One patient (female, aged 20) developed agranulocytosis 11 weeks after starting treatment and 6 weeks after beginning treatment with a daily dose of 105 mg/kg. This patient's neutrophil count recovered spontaneously 7 weeks after the discontinuation of L1. A decrease in serum zinc levels to subnormal levels was observed in four patients with symptoms of dry skin, with an itchy scaly rash in two that was associated with low serum zinc levels that responded to zinc therapy. Urinary zinc levels ranged from 4.7 to 23.4 (13 +/- 5.5) mumol/24 h and were above 9 mumol/24 h (upper limit of normal) in eight patients. Mild nausea occurred in three patients and transient diarrhea in a fourth. Mild musculoskeletal symptoms occurred in three patients but settled without discontinuation of L1 therapy in two and with temporary discontinuation of L1 in the third. A transient increase in serum aspartate transaminase was also noted in five patients, but serum aspartate transaminase levels subsequently decreased in all of them. No cardiovascular, neurologic, renal, or retinal toxicities were demonstrable. These results confirm that L1 is an effective oral iron chelator. Further clinical trials are needed to determine the incidence and severity of adverse effects.