Twenty-five previously untreated acne patients were monitored throughout a 6-month course of therapy with either tetracycline or minocycline for changes in the numbers of staphylococci, propionibacteria and yeasts of the genus Malessezia on the skin surface. Antibiotic resistant staphylococci and propionibacteria were also counted. Minocycline (50 mg b.d.) produced a 10-fold greater reduction in propionibacterial numbers compared to tetracycline (500 mg b.d.) after 12 (P < 0.02, t-test) and 24 weeks (P < 0.05) of therapy. As treatment progressed, propionibacteria were replaced by yeasts, numbers of which were significantly increased by week 12 (P < 0.02) in tetracycline-treated patients and by week 24 (P < 0.01) in minocycline-treated patients. This suggests that yeasts have no role in the pathogenesis of acne but may compete with propionibacteria for the same niche. Overgrowth of antibiotic resistant staphylococci prevented any decrease in staphylococcal numbers in tetracycline-treated patients, but minocycline produced a significant and sustained reduction in staphylococcal numbers after 1 week of therapy (P < 0.001). An increase in the number of multiply resistant (geqslant R: gt-or-equal, slanted 3 resistances) staphylococci occurred in 67% of tetracycline-treated and 33% of minocycline-treated patients by the end of the treatment period. There was no evidence of propionibacterial resistance in either treatment group. This study shows that minocycline has much greater antibacterial activity in vivo against both staphylococci and propionibacteria and produces less staphylococcal antibiotic resistance than tetracycline.