The efficiency of prophylactic antibiotic therapy in amputation surgery was studied in a prospective, randomized trial of a first-generation cephalosporin (cephalothin) compared with a narrow-spectrum beta-lactam stable penicillin (methicillin). Eighty-eight patients received cephalothin 2 g X 4 on the day of operation, while 86 patients received methicillin 1 g X 4. The patients were followed up for 21 days. Infected wounds occurred in 14.8% of the patients in the cephalothin group, compared with 14% in the methicillin group. The frequency of deep infections was 10.2% versus 4.7% (P = 0.1611). The reamputation frequency was 18.2% in the cephalothin group compared with 12.8% in the methicillin group; the frequency of below-knee reamputation was 18.4% versus 7.7% (P = 0.1469). No clostridial infections were found. The study did not demonstrate any significant difference between cephalothin and methicillin in the prophylaxis for lower-extremity amputations, although the latter drug tended to be the best choice.