Forty-three hemophiliacs with AIDS or ARC received a daily dose of 334 or 500 mg didanosine (2',3'-dideoxyinosine or ddI) orally in 2 divided doses in phase I/II, open-label clinical trial conducted in Japan. Twenty-eight patients completed 6 months of therapy. There was an increase in circulating CD4(+) cells in 19 valuable patients from 91 +/- 25 (mean +/- SE) at entry to 131 +/- 38 at 24 weeks of therapy P = 0.01; Wilcoxon signed rank). Fourteen of 37 patients met the criteria for CD4 rise >/= 50/mm3 rise or >/= 50% increase from entry values) for more than 4 consecutive weeks. Twenty patients were p24 positive at entry. Nine out of the 10 evaluable patients (90%) showed a decline in p24 antigen at weeks 20-24 (P = 0.02). Thirty-five patients had symptoms related to HIV-1 infection at entry. Twenty-seven patients reported improvements in constitutional symptoms during therapy. Nine patients presented with possible drug-related adverse effects, and didanosine was discontinued in 6 patients (one each with edema; abdominal pain with anorexia; hematuria with edema and rash; sense of abdominal distension with anorexia; diarrhea and abdominal pain; and irritability). One patient had a transient increase in serum amylase level to twice the upper limit of normal, but he continued to receive the drug. These data suggest that didanosine was generally well tolerated in hemophiliacs with AIDS or ARC, and its administration correlated with improvement in constitutional symptoms and laboratory findings. The adverse effects of didanosine seen in this population were moderate to mild, and no complications related to hemorrhagic diathesis were observed, although the relative risk of acute pancreatitis in this population (while not seen in the present study to date) requires more study.