The efficacy and safety of cefpirome was reviewed from the documentation of comparative pivotal trials in patients with urinary tract or lower respiratory tract infections UTIs and LRTIs, respectively). A majority of patients with UTIs had pyelonephritis and/or complicated UTIs. Most patients with LRTIs had community acquired pneumonia. Studies of UTI included 865 patients treated with cefpirome 1 g bid and 443 patients allocated to ceftazidime 1 g bid. Satisfactory clinical outcome was reported in 87% and 83%, respectively. Eradication of organisms causing bacteriuria was achieved in 87% and 86%, respectively. In the LRTI trials 199 patients received cefpirome 1 g bid and in 653 patients it was dosed 2 g bid. Comparators were ceftazidime 2 g bid (N = 197) or 2 g tid (N = 296) or ceftriaxone 1 g bid (N = 77). With all treatments unsatisfactory clinical or bacteriological outcome was recorded in < 15% of the patients. The safety of cefpirome and comparators was evaluated in pivotal phase II and III studies and deaths were analysed in all clinical trials for which data were available by June 30th 1991. Cefpirome did not differ from comparators in terms of frequencies or distribution within body systems of adverse events. Death rates were 3.9% in 9189 patients receiving cefpirome and 5.1% in 3162 receiving a comparator. The deaths were in an absolute majority of cases not considered related to study drug given. The most common cause of death was infection, indicating that the trial samples were selected from populations of patients with serious infections. Cefpirome was as safe and efficaceous as its comparators and is a new injectable cephalosporin with broader spectrum than ceftazidime. It should be a suitable alternative for empiric treatment of serious infections in hospitalised patients.