Serum levels of minocycline hydrochloride were determined by bioassay in a total of 223 acne patients (123 male, 100 female) receiving either the recommended dose (100 mg/day) or a high dose (200 mg/day) of the standard preparation (101 patients) or a modified release formulation (132 patients). Sera were collected within 6 h of the morning dose 7-10 days after the start of treatment. Mean minocycline serum levels were consistently higher in females than in males, irrespective of dose or formulation. The difference only reached statistical significance (P<0.05, Student's t-test) in the case of the standard preparation at a dose of 50 mg, b.d. Serum levels were increased significantly in both sexes at the higher dosage of each formulation (P<0.01) but there was no significant difference between formulations at either dosage. Variation in serum concentrations was not accounted for by variation in body mass. Serum levels above the modal minimum inhibitory concentration (MIC) of minocycline for fully sensitive strains of Propionibacterium acnes I (0.125 mu g/mL) were recorded in all patients. In contrast, serum levels equal to or greater than the modal MIC of minocycline for resistant propionibacteria (2 mu g/mL) were recorded in only 17.9% of patients on the low dose standard preparation compared with 55.6% on the high dose standard preparation (P<0.001, chi(2)). Even in females on the high-dose modified release formulation, 32.2% had serum levels below the modal MIC of minocycline for resistant strains. We conclude that, in terms of achievable serum levels over a short time period, there is no advantage of the modified release formulation over the standard preparation of minocycline. Whichever formulation is used, dose manipulation may be necessary to achieve maximum therapeutic benefit, especially in those individuals who are colonized by propionibacteria with reduced sensitivity to minocycline.