In an effort to isolate the effect of theophylline on sleep from the effect of asthma, we examined the impact of oral theophylline on sleep quality in normal, nonasthmatic subjects. Ten healthy, nonsmoking, male subjects ranging in age from 23 to 35 were studied. The subjects were randomly divided into two groups and studied in a double-blinded, crossover designed protocol. Each group underwent two consecutive nights in the sleep laboratory. Group one received four doses of 300 mg of sustained-release theophylline at 12-h intervals prior to the study night. Group two received a placebo. After 14 days, the patients crossed over to the other group, and the procotcol was repeated. The mean serum theophylline level on the morning following the study night was 7.85 +/- SD 2.12 micrograms/ml, with a range of 5.1 to 12.1 microns/ml. The analysis of variance for a crossover design showed no order effect. The analysis for drug effect showed that theophylline administration resulted in a statistically significant adverse effect on arousals per sleep hour (19.3 vs 15.9 for placebo, p = .006) and total sleep time (370.9 min vs 399.45 min for placebo, p = .015). Comparing the sum of ranks data, theophylline was found to have a significant adverse effect on sleep quality (p = .036). We conclude that low doses of oral theophylline result in a significant disturbance in sleep quality in normal nonasthmatic subjects.