A mouse model was developed to evaluate the efficacy of antibiotic treatment of pneumonic plague; streptomycin was compared to antibiotics with which there is little or no clinical experience. Infection was induced by inhalation of aerosolized Yersinia pestis organisms. Antibiotics were administered by intraperitoneal injection every 6 hours for 5 days, at doses that produced levels of drug in serum comparable to those observed in humans treated for other serious infections. These studies compared in vitro to in vivo activity and evaluated the efficacy of antibiotics started at different times after exposure. Early treatment (started 24 h after challenge, when 0 of 10 mice tested had positive blood cultures) with netilmicin, ciprofloxacin, ofloxacin, ceftriaxone, ceftazidime, aztreonam, ampicillin, and rifampin (but not cefazolin, cefotetan, or ceftizoxime) demonstrated efficacy comparable to streptomycin. Late treatment (started 42 h after exposure, when five of five mice tested had positive blood cultures) with netilmicin, ciprofloxacin, ofloxacin, and a high dose (20 mg/kg of body weight every 6 h) of gentamicin produced survival rates comparable to that with streptomycin, while all of the beta-lactam antibiotics (cefazolin, cefotetan, ceftriaxone, ceftazidime, aztreonam, and ampicillin) and rifampin were significantly inferior to streptomycin. In fact, all groups of mice treated late with beta-lactam antibiotics experienced accelerated mortality rates compared to normal-saline-treated control mice. These studies indicate that netilmicin, gentamicin, ciprofloxacin, and ofloxacin may be alternatives for the treatment of pneumonic plague in humans. However, the beta-lactam antibiotics are not recommended, based upon poor efficacy in this mouse model of pneumonic plague, particularly when pneumonic plague may be associated with bacteremia.