Parasympathetic blockade shortens the duration of the Q-T interval and ventricular effective refractory period independent of heart rate change. Since relative parasympathetic effect increases during sleep, it was determined whether sleep was associated with a change in the Q-T interval. Fifteen patients receiving no drugs underwent 3 to 6 days of continuous electrocardiographic recordings. Tracings were sampled every 30 minutes and recorded at a paper speed of 25 mm/s. This provided 12,000 Q-T and R-R intervals that were measured. Comparison of R-R intervals that had similar durations during sleep and awake states revealed that the duration of the Q-T interval was longer during sleep in all 15 patients (p less than 0.001). Eight patients had sufficient range of overlap of R-R intervals to compare linear regression lines of Q-T intervals recorded while awake with Q-T intervals recorded while asleep. The regression lines during sleep exhibited a mean intercept change of 38 +/- 37 ms and mean slope change of -0.021 +/- 0.040 ms when compared with the regression lines during the awake state. The difference in Q-T interval between awake and sleep states was 19 +/- 7 ms when calculated at a heart rate of 60 beats/min. These statistical comparisons of the relationship of the Q-T interval to R-R interval indicate that the Q-T interval is longer during sleep than during the awake state at the same heart rate. Prolongation of the Q-T interval during sleep may reflect increased vagal tone or sympathetic withdrawal. These changes in repolarization may be related to the diurnal variation of some ventricular arrhythmias.