Broad spectrum cephalosporins have been studied extensively in animal models of experimental infections. There is generally good correlation between the results of therapy of experimental infections and clinical trials in humans. However, the results of animal model studies are better predictors of the failure than of the success of a chemotherapeutic regimen. Cefotaxime and the new 'fourth' generation agent, cefpirome, were comparable in the treatment of experimental meningitis caused by Streptococcus pneumoniae. Cefpirome was the most effective cephalosporin as therapy for methicillin-susceptible Staphylococcus aureus experimental endocarditis. The most effective broad spectrum cephalosporins for the treatment of Gram-negative experimental pneumonia were cefpirome, cefotaxime and cefodizime. Cefpirome was equivalent to ceftazidime or cefazolin as treatment for Pseudomonas aeruginosa or methicillin-susceptible S. aureus experimental osteomyelitis. Because of its potent activity in vitro and in animal models of experimental infections caused by methicillin-susceptible S. aureus and Gram-negative bacilli, cefpirome may offer a therapeutic advantage over currently available broad spectrum cephalosporins.