Glyburide (glibenclamide) is a specific blocker of the adenosine triphosphate (ATP) sensitive potassium (K+) channel. It has been reported to result in prolongation of the QT interval. QT interval dispersion (QTd) is a potentially sensitive marker for increased risk of arrhythmia and sudden cardiac death. The aim of the present study was to evaluate the effect of glyburide on QTd and compare it with that of metformin, a hypoglycemic agent that does not block the adenosine triphosphate sensitive K+ channel. Thirty patients with type 2 diabetes were randomized to glyburide and metformin groups. A 12-lead electrocardiogram was obtained before and at 2 months after being on glyburide or metformin. Therapy with QT and QTd were measured and QT corrected for rate (QTc). There was no significant difference between the glyburide and metformin groups in age (62 +/- 9 vs 59 +/- 10 years), baseline RR interval (819 +/- 86 vs 753 +/- 100 ms), QT (387 +/- 28 vs 383 +/- 27 ms), and QTc (433 +/- 25 vs 444 +/- 15 ms). Glyburide was associated with a significant increase in QTc (433 +/- 24 to 467 +/- 24 ms, p <0.001), QTd (24 +/- 16 to 60 +/- 22 ms, p <0.001), and QTc dispersion (QTcd) (35 +/- 18 to 68 +/- 21 ms, p <0.001). In contrast, metformin was associated with a decrease in QTc (444 +/- 15 to 432 +/- 15 ms, p <0.01) and did not affect QTd (14 +/- 5 to 12 +/- 6 ms, p = NS) and QTcd (23 +/- 9 to 22 +/- 10 ms, p = NS). Glyburide, unlike metformin, causes an increase in QT dispersion. Increased dispersion may be a factor underlying an increased risk of arrhythmias and sudden cardiac death.