The effects of streptozotocin-induced diabetes (5-7 days or 7 wk) on cardiac ATP-sensitive potassium channels (KATP channels) were investigated with the use of single-channel and action potential recordings from dissociated ventricular myocytes isolated from control and diabetic rat hearts. In inside-out patches from diabetic myocytes (5-7 days), the IC50 for ATP inhibition was 82 +/- 7.2 microM (mean +/- SE, n = 8), twice that in controls (43 +/- 3.6 microM, n = 12). For 7-wk diabetic rats, the IC50 was 75 +/- 2.3 microM (n = 6). Increasing internal ADP concentration attenuated ATP-induced inhibition in both controls and diabetics. On reducing the internal pH from 7.4 to 6.8, both control and diabetic myocytes showed a 1.7-fold increase in the IC50 for ATP inhibition. No differences were observed in either intraburst kinetics or unitary conductance of single channels from control and diabetic myocytes. In diabetic myocytes, action potential duration at 90% repolarization (APD90) was longer and more variable than in controls and was significantly shortened by application of the KATP channel opener cromakalim (50 microM). Cromakalim scarcely affected APD90 in controls. Computer simulation of the longer diabetic APD90 required a lower background conductance during the plateau phase in addition to small, measured changes in the delayed rectifier current, transient outward current, and ATP-sensitive K+ current (KATP current, IKATP). The simulations reproduced the enhanced sensitivity of the diabetic APD90 to changes in IKATP. These results have important implications for cardiac function in diabetics and their treatment by sulfonylureas.