The effectiveness of topical phenytoin as a wound healing agent was compared with that of OpSite (Smith & Nephew) and a conventional topical antibiotic dressing (Soframycin, Roussel) in a controlled study of 60 patients with partial-thickness skin autograft donor sites on the lower extremities. Mean time to complete healing (complete epithelialization) was 6.2 +/- 1.6 days in the phenytoin-treated group (30 patients), compared to 8.6 +/- 2.2 days with OpSite (15 patients), and 12.6 +/- 3.4 days in the 15 Soframycin-treated patients. The differences between the treatment groups were significant at P < 0.001. Mean pain scores were also lower in the phenytoin-treated group, 0.40 +/- 0.55 vs. 0.66 +/- 0.60 with OpSite (P < 0.05) and 1.4 +/- 0.50 with the conventional dressing (P < 0.001). Both phenytoin and OpSite were superior to the Soframycin dressing with respect to bacterial contamination and wound infection as measured by Gram stains of wound smears, swab and aspirate (OpSite) cultures, and clinical assessments (P < 0.001) carried out on the fifth day of treatment. No local or systemic adverse effects of the three agents used were noted. Phenytoin appears to be an effective, low-cost and safe method for the treatment of partial-thickness skin graft donor sites, comparing very favourably with, and in some aspects superior to, occlusive dressings. Further clinical use and evaluation of topical phenytoin are merited.