Seventy-one adult patients with 72 infections were treated, by random selection, with intravenous/oral ciprofloxacin or intravenously administered ceftazidime. Twenty-seven additional patients with 29 infections who were not appropriate for random assignment were treated in an open study with intravenously administered ciprofloxacin only; the latter infections were generally more serious or were caused by ceftazidime-resistant organisms. The most common doses were ciprofloxacin, 200 mg intravenously and 500 mg orally every 12 hours and ceftazidime, 1 to 2 g intravenously every eight to 12 hours. Forty-seven ciprofloxacin-treated infections and 31 ceftazidime-treated infections were evaluable for determination of efficacy. Infections included lower respiratory tract (21 infections), urinary (37 infections), skin/soft tissue (14 infections), bacteremia/endocarditis (four infections), colitis (one infection), and mastoiditis (one infection). Median minimal inhibitory concentrations of ciprofloxacin and ceftazidime were, respectively: for Enterobacteriaceae, Haemophilus influenzae, and Branhamella catarrhalis, no more than 0.06 and no more than 0.25 micrograms/ml; for Pseudomonas aeruginosa, 0.25 and 4 micrograms/ml; for Enterococcus faecalis, 1 and more than 32 micrograms/ml; and for Staphylococcus aureus, 0.25 and 8 micrograms/ml. Ciprofloxacin, 200 mg intravenously, yielded mean serum concentrations 0.5 and eight hours post-intravenous infusion of 2.3 and 0.7 micrograms/ml, respectively. Satisfactory clinical responses were achieved in 17 (81 percent) of 21 patients with intravenous/oral ciprofloxacin, 22 (71 percent) of 31 patients with ceftazidime, and 20 (77 percent) of 26 patients with intravenous ciprofloxacin. The most common treatment failures occurred in complicated skin/soft-tissue infections treated with intravenous/oral ciprofloxacin, complicated urinary tract infections treated with ceftazidime, and necrotizing P. aeruginosa pneumonia treated with intravenous ciprofloxacin; the pneumonia patients all had respiratory failure and had been previously unresponsive to treatment with other appropriate drugs. Serious adverse reactions were observed in three patients, seizures with intravenous ciprofloxacin in two patients, and Clostridium difficile diarrhea with ceftazidime in one patient. We conclude that sequential intravenous/oral ciprofloxacin and ceftazidime were comparable in efficacy and safety; the ability to change from intravenous to oral therapy is a major convenience. Intravenous ciprofloxacin was useful for more serious infections, often caused by ceftazidime-resistant organisms.