A study was carried out in 90 patients with superficial, soft-tissue infections to compare the effectiveness of topical treatment with 2% sodium fusidate ointment used alone, and oral antibiotic therapy (clindamycin, erythromycin, or flucloxacillin) plus a placebo ointment. The results showed that the number of days for healing to take place was significantly shorter with sodium fusidate ointment, and there was also a highly significant preference for it over oral antibiotic therapy in the subjective assessment of clinical response. Bacteriological investigations of swabs from 58 of the patients showed Staphylococcus aureus to be the most frequently isolated pathogen: 72% of the strains were penicillin-resistant but all were sensitive to sodium fusidate. It is suggested that oral antibiotic therapy should be reserved for those cases where there is evidence of systemic spread of the infection and that sodium fusidate ointment, with or without surgical drainage, should be the standard initial treatment in the out-patient and general practice situation.