The treatment of opportunistic infections in AIDS patients is changing rapidly as new drugs become available and new studies of old drugs are completed. I have tried to use the AIDS commentaries to keep up with these advances. Since fluconazole has just recently been approved by the US Food and Drug Administration, I asked Dr. Robert A. Larsen, an active clinical investigator at the University of Southern California School of Medicine, to review his current recommendations for the use of the azoles in AIDS patients. The response to therapy of the various mycoses in patients with advanced human immunodeficiency virus (HIV) infection is dramatically different from that in other patient populations and has different end points in each group. Cure is uncommon in HIV-infected patients, and relapsing infection when antifungal therapy is stopped is the rule. Control of infection with relief of symptoms and return to productive, high-quality life is therefore a commendable goal and a reasonable end point. Thus, the azoles, especially fluconazole, are important additions. From the data presented here it appears that amphotericin may still have an edge over fluconazole for acute therapy of cryptococcal meningitis in sicker patients, at least for the first several weeks (although fluconazole may be as good and is certainly less toxic and easier to administer for patients who can take oral medications). Whether 5FC should be added when amphotericin is used for acute therapy is still controversial. Dr. Larsen appears to favor it while my group feels that the potentiation of bone marrow toxicity without any clear evidence of enhanced efficacy in AIDS patients argues against its use.(ABSTRACT TRUNCATED AT 250 WORDS)