BACKGROUND Topical treatment of acne vulgaris on the back is challenging largely due to the extensive broad surface with difficult to reach areas. A "leave-on" foam is suited for application to the trunk due to ease of application and spreadability. Prior to this trial, no data on Propionibacterium acnes (P. acnes) reduction on the back has existed for any benzoyl peroxide (BP) formulations or other acne treatments. OBJECTIVE To evaluate the effectiveness of BP (5.3%) emollient foam and BP (8%) wash in reducing P. acnes levels on the back. METHODS Five-week open-label single-center study of 20 healthy subjects (> 18 years old), colonized with P. acnes on their backs (> 10,000 colonies per cm2). Subjects were treated once daily with BP (5.3%) foam for two weeks; no treatment in week 3, and BP (8%) wash once daily for two further weeks. Quantitative bacteriologic cultures obtained at baseline and weeks 1, 2, 3, 5 and 6. RESULTS Nineteen evaluable patients. Total P. acnes counts were reduced by 1.9 log (one week) and 2.1 log (two weeks) with BP (5.3%) emollient foam. BP (8%) wash did not reduce P. acnes counts after two weeks. CONCLUSIONS BP (5.3%) emollient foam was superior to BP (8%) wash in reducing P. acnes on the back. The lack of effect of BP (8%) wash is surprising in view of the demonstrated results on the face and warrants further study.