OBJECTIVE To investigate the cardiac effect of QT interval prolongation in the treatment of acute promyelocytic leukemia (APL) with arsenic trioxide (As(2)O(3)), and the relationship between QT and serum arsenic concentration. METHODS Blood serum arsenic concentrations of thirty APL patients were determined at 2 hours, 4 hours, 8 hours, and 24 hours after As(2)O(3) injection using atomic fluorophotometry. Cardiac functions were measured simultaneously using a 12-lead body-surface electrocardiogram (ECG). Q-T intervals were manually measured, and then corrected using Bazett's formula (QTc). QT dispersion (QTd) was also calculated. In order to assess the effects of arsenic on the symptoms of anemia, twenty-four anemia patients were divided into two groups on the basis hemoglobin concentration: Group 1 (Hb > or = 90 g/L), and Group 2 (60 g/L < or = Hb < 90 g/L). QTc and QTd of these patients were also manually measured. RESULTS All QT intervals of APL patients treated with As(2)O(3) injection were prolonged [32.2 ms (27, 41 ms); P < 0.05], but the changes of QTd were not prominent [3 ms (-8, 7 ms), P > 0.05]. There was a delay of 2 hours in maximum QTc following peaks in serum arsenic concentration. Changes in QTc and QTd of the two anemic groups were not prominent. CONCLUSIONS As(2)O(3) can prolong QTc intervals in APL patients, but the effects are delayed compared to peak serum arsenic concentrations. As(2)O(3) has no prolongation effect on QTd. Mild and moderate anemia do not effect QTc and QTd.