This study was designed to investigate the interaction between high-dose oral ganciclovir (6,000 mg/day) and didanosine at steady state in patients who were seropositive for human immunodeficiency virus (HIV) and cytomegalovirus (CMV) infection. The study was conducted as an open-label, randomized, three-period crossover study. Patients received (in random order) multiple oral doses of didanosine 200 mg every 12 hours alone, ganciclovir 2,000 mg every 8 hours alone, and ganciclovir 2,000 mg every 8 hours in combination with didanosine 200 mg every 12 hours. Blood and urine samples for determinations of drug concentrations were obtained on day 3 of each dose regimen. When ganciclovir was administered either before or 2 hours after didanosine, the mean increases in maximum concentration (Cmax), area under the concentration-time curve (AUC0-12), and percent excreted in urine of didanosine were 58.6% and 87.3%, 87.3% and 124%, and 100% and 153%, respectively. There were no statistically significant effects of didanosine on the steady-state pharmacokinetics of ganciclovir in the presence of didanosine, irrespective of sequence of administration. There were no significant changes in renal clearance of didanosine, suggesting that the mechanism for the interaction does not involve competition for active renal tubular secretion. The mechanism responsible for increased didanosine concentrations and percent excreted in urine during concurrent ganciclovir therapy may be a result of increased bioavailability of didanosine. However, the mechanism appears to be saturated at oral ganciclovir doses of 3 g/day.