Our objective was to define the actions of sematilide in rabbits and to assess the contribution of the delayed rectifier (IK) to rate dependence of action potential duration (APD) in rabbit ventricular myocardium. In studies in vivo, New Zealand White rabbits were used to obtain dose/response curves of the effects of sematilide on APD from contact monophasic action potentials (MAP), ventricular effective refractory period (VERP), and ECG. Sematilide or placebo was administered as an i.v. bolus followed by a 45-min infusion in the following cumulative manner: infusion 1 (1 mg/kg bolus + 8 micrograms/kg/min); infusion 2 (2 mg/kg + 20 micrograms/kg/min); and infusion 3 (7 mg/kg + 68 micrograms/kg/min). At each infusion level, VERP and APD at 75% repolarization (APD75) were measured during cardiac pacing between 200- and 400-ms cycle length (CL). Serum sematilide levels were analyzed by high-performance liquid chromatography (HPLC). In studies in vitro, sematilide's effects on the delayed rectifier were assessed in isolated rabbit ventricular myocytes by using patch-clamp techniques. Sematilide infusion in vivo resulted in stable serum levels of 1.3 +/- 0.5, 3.7 +/- 1.4, and 13.4 +/- 1.8 micrograms/ml during infusions 1, 2, and 3, respectively. Maximal effects occurred at infusion 2, such that at 400 ms CL, sematilide widened predrug APD75 (145 +/- 5 ms) by 27 +/- 4% (p < 0.001 vs. placebo), and at 200-ms CL, sematilide prolonged predrug APD75 (115 +/- 10 ms) by only 18 +/- 4% (p < 0.001 vs. placebo; p < 0.05 vs. 400-ms CL). Similar effects were observed in VERP. Sematilide enhanced rate dependence of APD and produced the same degree of APD prolongation at a given CL, during accommodation to and recovery from rapid pacing. Rabbit ventricular myocytes appeared to have at least two types of delayed rectifier. Sematilide selectively blocked IKr, and block was not relieved by repetitive stimulation. In conclusion, the APD-widening effect of sematilide was independent of previous pacing history. Sematilide had little influence on background processes likely responsible for shortening APD because of rapid repetitive stimulation.