In the summer of 1988, dose-finding phase 1 trials of the efficacy and safety of didanosine were begun by investigators at the National Cancer Institute (NCI), by the AIDS Clinical Trials Group (ACTG) at New York University School of Medicine and the University of Rochester School of Medicine, and by researchers at Boston City Hospital (BCH). The schedules of drug administration studied included once daily (BCH), twice daily (ACTG and NCI), and three times daily (NCI). The total daily dose studied ranged from 0.8 to 66 mg/(kg.d). Significant toxicities developed in more than half of the patients receiving doses of > or = 12 mg/(kg.d). Only 5% of patients given < 10 mg/(kg.d) developed grade 3 or 4 toxicities. Although CD4+ lymphocyte counts increased at doses as low as 1.6 mg/(kg.d), a minimal effective dose could not be determined. An analysis of data on serum levels of p24 antigen suggests that a dose of > or = 5 mg/(kg.d) is superior to lower doses in terms of activity against human immunodeficiency virus. A dose given on a once-daily schedule may exhibit less antiretroviral activity than the same dose given on a twice-daily schedule. Long-term follow-up of the phase 1 trial of the once-daily regimen at BCH has further documented the safety of didanosine given at doses of < 12 mg/(kg.d) for extended periods. A review of these studies suggests that a twice-daily schedule of didanosine administration at a daily dose of 5-10 mg/(kg.d) is safe and is associated with significant antiretroviral activity in vivo.