This study was designed to establish in vivo the possible presence of antimicrobial substance(s) in human skin. If such agents are present on the skin surface, their removal might allow increased multiplication of applied organisms. One forearm was washed repeatedly for 5 minutes with lipid solvents; the other served as the control. The organisms applied (3 × 10 3 -7 × 10 4 in 0.01 ml Ringer's solution) on the forearms were not allowed to dry and were covered immediately with plastic protective cups for 5 hours. The microorganisms were harvested using a modified Pachtman-Burtenshaw method. With Staphylococcus aureus and Streptococcus pyogenes , acetone washed forearms had 2- to 510-fold more organisms than the control arm (P = 0.0008 and 0.08, respectively). Similar results were noted with ether (average 1:62, p=0.005). Candida albicans increased 2- to 200-fold (average 1:37, p=0.002). This effect did not occur with Escherichia coli and Pseudomonas aeruginosa (P = 0.8). To determine the time required for natural replenishment of the antimicrobial substances, bacteria were applied 2, 3, and 5 hours after washing with acetone. Increased numbers were seen at 2 (P = 0.04), but not at 5 (P = 0.2) hours. On the contrary, microbial flora removed by acetone washing were not restored during the 5 hour period. Regional variation studies showed that the forehead did not inhibit S. aureus persistence (as did the forearms) (P = 0.003). Forearms treated with acetone and the extracts not removed did not show increase of S. aureus counts when compared with untreated arms (average ratio 1 : 1, p=0.6). When acetone extracts of the skin surface were removed, evaporated, and then replaced on the skin, the increased bacterial counts noted above were no longer found.