Twice-daily intramuscular ceforanide therapy of Staphylococcus aureus endocarditis in parenteral drug abusers was compared in a randomized prospective trial with intravenous cephapirin therapy. Dosage regimens were ceforanide, 1 g every 12 h, and cephapirin, 2 g every 4 h. Mean minimal inhibitory and bactericidal concentrations of ceforanide for S. aureus treated with ceforanide were 0.78 and 1.56 microgram/ml compared to cephapirin for patient isolates of 0.08 and 0.14 microgram/ml, respectively. Serum killing levels with ceforanide were 1:5.7 and 1:1.5 at peak and trough levels, compared to 1:134 (peak) and 1:4.2 (trough) with cephapirin. Despite this apparent in vitro advantage of cephapirin, patients treated with ceforanide did as well as those with cephapirin. Of 16 ceforanide-treated patients, all responded initially to therapy, and 15 were cured with 28 days of therapy. One patient relapsed at the end of therapy. Of 16 cephapirin-treated patients, 1 was a clinical and microbiological failure, and 3 other relapsed at the end of therapy. In addition, one ceforanide-treated patient and two cephapirin-treated patients developed central nervous system abscesses. These were cured with drainage and continuation of antibiotic therapy. Ceforanide was well tolerated by the intramuscular route. Cost analysis suggests that therapy with intramuscular ceforanide would result in an approximate 70% decrease in drug therapy cost when compared to intravenous cephapirin. Ceforanide appears to be a safe, efficacious, convenient, and relatively inexpensive drug for treating staphylococcal endocarditis in parenteral drug abusers.