A randomized, double-blind trial compared the clinical and bacteriologic efficacy of ampicillin/sulbactam (2 g/1 g) and cefoxitin (2 g) administered intravenously every 6 h to patients with (n=49) or without (n=47) histories of injection drug abuse who presented with cutaneous or other soft-tissue infections. Cure or improvement occurred in 89.8% of ampicillin/sulbactam-treated patients, compared with 93.6% of cefoxitin-treated patients. The median time to resolution of all symptoms was 10.5 days with ampicillin/sulbactam treatment and 15.5 days with cefoxitin treatment. Mixed aerobic-anaerobic infection was encountered frequently in both treatment groups. A significantly higher percentage of Streptococcus species was found in the major abscesses of the patients with histories of injection drug abuse, compared with those without such histories (37% vs. 19%, respectively; P=.0009). Overall, ampicillin/sulbactam eradicated pathogens from the major abscesses in 100% of patients, whereas the eradication rate with cefoxitin was 97.9%. The 2 drugs were well tolerated. Ampicillin/sulbactam and cefoxitin were equally effective for the empirical treatment of cutaneous or other soft-tissue infections in injection drug abusers and patients who did not inject drugs.