A prospective, randomized trial was performed to compare the incidence of wound infection following skin closure with subcuticular polyglycolic acid and interrupted monofilament nylon mattress sutures in patients undergoing high saphenous ligation (126 patients; 75 polyglycolic acid wounds, 86 interrupted monofilament nylon mattress wounds). By 6 weeks after operation, pus had discharged from 15 (20%) polyglycolic acid wounds compared with seven (8%) interrupted monofilament nylon mattress wounds (P = 0.05), giving an overall infection rate of 13.7%. Infection occurred in 23% of the 56 procedures performed by the two surgeons with the highest infection rates compared with 8% of the 105 procedures performed by the remaining nine surgeons (P = 0.02). The excess of infections in wounds closed with subcuticular polyglycolic acid was caused by an excess of infections in operations performed by the two surgeons with the highest infection rates. The higher infection rate found with subcuticular polyglycolic acid appeared to be operator dependent.