From November 1979 to July 1986, 52 patients (27 women and 25 men; median age 52 years) with advanced adrenocortical carcinoma entered a prospective, nonrandomized study evaluating moderate-dose mitotane and doxorubicin hydrochloride (Adriamycin). Thirty-two tumors (62%) were well differentiated and evidence of hormone production was present in 24 patients (46%). Patients with well-differentiated or functional tumors received mitotane, 6 gm daily; patients for whom mitotane failed or those with poorly differentiated, non-hormone-producing tumors received Adriamycin, 60 mg/m2 every 3 weeks. Initially, 36 patients were treated with mitotane and 16 patients with Adriamycin. Eight patients (22%) responded to mitotane and three (19%) responded to Adriamycin. No response was noted in the 15 patients for whom mitotane failed and who then received Adriamycin. Severe toxicity occurred in 36% of patients who received mitotane and in 26% who received Adriamycin. Overall median survival after onset of treatment was 14 months. We conclude that mitotane or Adriamycin used initially can induce tumor regression in about 22% and 19% of selected patients, respectively. However, Adriamycin is ineffective as second-line chemotherapy for patients with well-differentiated or functioning tumors for whom mitotane is ineffective.