There is evidence to suggest that single-agent broad spectrum antibacterials may be cost-effective alternatives to combination antibiotics for the empirical management of febrile neutropenia in cancer patients. The objectives of the present study were 2-fold. The first objective was to compare the clinical effectiveness of ceftazidime monotherapy with that of 2 combination antibiotic regimens in cancer patients with febrile neutropenia. The 2 comparator regimens consisted of tobramycin plus piperacillin, either with (regimen 'CAP') or without (regimen 'AP') cefazolin. The second objective was to perform a cost-effectiveness analysis of the 3 regimens. Meta-analysis of randomised comparative trials between the 3 therapy groups was performed to determine the average overall response rate after 3 to 5 days of treatment. Seven clinical studies were selected for analysis. The overall incidence of adverse drug reactions (ADRs) was determined using the results of comparative and noncomparative studies. A comparative cost-analytic model was applied from a hospital perspective. The costs of primary therapy, hospitalisation, laboratory tests, routine patient care and treating ADRs were calculated, as were future costs. Monotherapy with ceftazidime was associated with an overall response rate of 63.5% and mean per-patient costs of $Can12,000 to $Can14,000. In comparison, regimen AP was associated with an overall response rate of 58.8% and mean costs of $Can13,000 to $Can16,000 per patient. The overall response rate in patients receiving CAP was 75.3%, and the mean cost per patient was $Can11,000 to $Can12,000. Thus, regimen CAP was the most cost-effective therapy from a hospital perspective.