Thirty dogs with superficial pyoderma were randomly allocated to treatments with either lincomycin hydrochloride (22 mg/kg twice daily) or clindamycin hydrochloride (11 mg/kg once daily), initially for three weeks. Samples were taken from pustules, from adjacent apparently uninvolved skin, and from the nares. These were submitted for bacterial culture and sensitivity testing. The dogs were re-examined after three weeks treatment and samples for bacteriology were taken from the nares, from any pustules that were present or from skin in the area that was previously affected; the treatment was extended if necessary. Seventy-one per cent of the dogs given lincomycin hydrochloride responded within three weeks compared with 81 per cent of the dogs treated with clindamycin hydrochloride. The overall response rates, including those given longer courses of treatment were 93 per cent for those treated with lincomycin hydrochloride and 94 per cent for those treated with clindamycin hydrochloride, and there was no statistically significant difference between the groups either after three weeks treatment or after extended treatment. The microbiological results demonstrated that Staphylococcus intermedius was present on the skin adjacent to pustules and suggested that the nasal carriage of S intermedius was a result of cutaneous colonisation.