Moxalactam was administered (20 mg/kg intravenously every 8 hours) as single-drug empiric antimicrobial therapy to 63 patients with bacteremia who were neither neutropenic nor immunosuppressed. Six patients (10%) had microorganisms that were susceptible to moxalactam and resistant to all other antimicrobial agents tested; two patients (3%) had microorganisms that were resistant to moxalactam and other agents tested. Of these 63 patients, 47 (75%) were cured with moxalactam therapy. Nine patients (14%) had breakthrough bacteremia while receiving other antimicrobial therapy and were cured subsequently with moxalactam therapy alone. The two major risk factors for failure of moxalactam therapy were polymicrobial bacteremia and an extrahepatic intra-abdominal source of infection; these two conditions frequently coexisted. Six of nine patients with polymicrobial bacteremia died. Superinfection (one pseudomonal, five enterococcal) was responsible for 6 of the 16 treatment failures. Enterococcal superinfection occurred exclusively among patients who had received relatively prolonged therapy with moxalactam for extrahepatic intra-abdominal infection, especially intraabdominal abscess. These five patients died, and postmortem examination showed that enterococcal superinfection was the major cause of death in all. Mild, reversible adverse reactions associated with use of moxalactam occurred in 14 of the 63 patients (22%). None had clinically overt bleeding. The use of moxalactam alone seems to be safe and effective and a cost-effective alternative empiric antimicrobial therapy for most patients with bacteremia who are not immunosuppressed or neutropenic and who are not at high risk of having Pseudomonas or polymicrobial bacteremia.