In two multicenter trials, a total of 576 patients with acute myeloid leukemia (AML) were treated and found to be evaluable. Two hundred forty-two patients were in a 1978 pilot study and 334 patients were in a 1982 randomized study. Ages were between 15 and 78 years (median, 48). The uniform remission induction therapy in both studies consisted of one to two courses of a 9-day combination of 6-thioguanine (TG) with cytosine arabinoside (ARA-C) and daunorubicin (DNR) [TAD9]. The timing and sequencing of TAD9 was designed according to cell kinetic effects of ARA-C. A complete remission (CR) was achieved in 65% (70% and 61%, respectively) of patients within a median of 33 days, and in 68% of responders after only one course. The CR rate in patients 60 to 78 years of age was 51% (66% and 39%, respectively). In the 1978 pilot study, different protocols of post-remission treatment were applied at the different centers: monthly 5-day maintenance, TAD9 consolidation, both consolidation and maintenance, or no further therapy. The group receiving treatment during CR showed 24% probability of remissions at 4 years v 0% probability of remissions in the untreated group. Between the different post-remission protocols, no significant differences were observed. Remission duration was not influenced by age, WBC, or morphologic cell type, but was longer in patients achieving CR within 30 days (P = .017). In the subsequent 1982 study, 145 patients in CR were randomized for TAD9 consolidation with or without monthly maintenance. The updated life-table analysis revealed a predicted rate of continuous remission at 2 1/2 years of 30% for the maintenance and 17% for the nonmaintenance arm (P = .003). These results of response and remission duration in adult patients of all ages support the validity of intensified induction therapy and of consequent myelosuppressive treatment in remission.