Clinical trial results have been analysed from 2607 patients (2953 infections, of which 38% were severe) treated with ceftazidime alone and from 466 patients (507 infections - 24% severe) treated with one or more comparative antibiotics. Of 235 staphylococcal infections (all sites) treated with ceftazidime alone 84% cleared and 16% relapsed or failed; of 28 infections treated with other antibiotics, 82% cleared and 18% relapsed or failed. Of 541 pseudomonas infections (all sites) treated with ceftazidime 60% cleared and 40% relapsed or failed; of 76 infections treated with other agents 43% cleared and 57% relapsed or failed. Adverse events were seen in 8.9% of the 2,607 ceftazidime treated and in 8.2% of the 466 patients treated with other agents and included local and gastrointestinal intolerance and hypersensitivity reactions. Changes in laboratory tests included eosinophilia, Coombs' positivity and liver enzyme increases. Superinfection occurred in 2.5% of patients treated with ceftazidime and in 5.2% of patients treated with other antibiotics. Deaths from all causes during or shortly after treatment were reported in 3.3% of the 2607 ceftazidime-treated and in 5.8% of the 466 patients treated with other agents. Factors, including neutropenia, influencing these differences are analysed.