Two-hundred-and-eight acne vulgaris patients were enrolled in a 24-week study to determine the bacterial resistance issues associated with the use of a topical 2% erythromycin gel. It consisted of a 12-week randomized, double-blind, parallel-group treatment phase comparing the active gel versus its vehicle followed by a 12-week single-blind regression phase with gel vehicle only. Bacteriological samples were taken from the face, back and nares for quantification by species and antibiotic resistance characteristics. Acne efficacy was assessed through week 12. The prevalence of erythromycin-resistant coagulase-negative Staphylococci on the face was extremely high (87%) at baseline, increased to 98% by week 12 in the erythromycin-treated group and did not change during regression. The density of these resistant organisms also significantly increased with erythromycin treatment with no change during regression. Similar prevalence and density patterns were also observed on the untreated back and in the nares. Nearly all of the resistant isolates were highly resistant (minimal inhibitory concentrations > 128 microg/ml). Resistance development was confined to the macrolide class of antibiotics. No anti-acne efficacy was observed.