The efficacy of clavulanate-potentiated amoxycillin was compared with amoxycillin alone in experimental staphylococcal infection in dogs and in a controlled trial in clinical cases of skin infection in dogs and cats. The experimental infection was produced by subdermal inoculation with beta-lactamase producing (amoxycillin resistant) staphylococci absorbed in cotton dust. This produced discrete, localised lesions with no systemic involvement. In a cross over study, six animals were randomly allocated to treatment with either amoxycillin alone (10 mg/kg, dosed twice daily) or a formulation of clavulanate-potentiated amoxycillin (12.5 mg/kg, of a 1:4 ratio, dosed twice daily). The lesions of the animals treated with clavulanate-potentiated amoxycillin resolved more quickly than those treated with amoxycillin alone. The difference was significant (P less than 0.05) for both lesion diameter and inflammation score after day 6 of treatment. A trial was carried out in clinical cases of skin disease which were randomly allocated to twice daily treatment with either amoxycillin alone (10 or 20 mg/kg), or with clavulanate-potentiated amoxycillin (12.5 or 25 mg/kg of a 1:4 ratio). The required duration of treatment was shorter (P less than 0.5) for the potentiated amoxycillin treatments, and the success rate (judged by cure or substantial improvement) was higher (P less than 0.05) for this group, especially (P less than 0.01) where amoxycillin resistant organisms were isolated. It was concluded that clavulanate-potentiated amoxycillin was an effective treatment of skin infections both under experimental and clinical conditions.