In a multicentre, prospective, controlled trial 211 patients with suspected septicaemia or pneumonia were allocated at random to either imipenem-cilastatin 500 mg 8-hourly or cefotaxime 1 g 6-hourly combined with amikacin 5 mg/kg 8-hourly. The treatments were administered for at least 5 days. Seventy patients on imipenem and 70 patients on cefotaxime-amikacin were assessable for comparison. There were no statistically significant differences between the two groups in underlying pathology and in the clinical results obtained: septicaemia 20/26 patients of the imipenem group and 20/25 patients of the cefotaxime-amikacin group were cured; pneumonia 38/44 patients of the imipenem group and 34/45 patients of the cefotaxime-amikacin group were cured. There were also no differences in the initial organisms and in the bacteriological cure rate, except for Pseudomonas aeruginosa. At the moment, imipenem administered alone is as effective as the cefotaxime-amikacin combination in the treatment of septicaemia or pneumonia in intensive care patients, with the exception of P. aeruginosa pneumonia in patients under assisted ventilation.