The current health-care philosophy dictates that new therapies should always be evaluated for their economic impact. Along with acquisition cost, the cost of delivery, monitoring, adverse effects and treatment failure must also be considered when determining the total cost of therapy. OBJECTIVE: The purpose of this study was to evaluate the efficacy, safety and cost of vancomycin and teicoplanin therapy in patients with neutropenia, after the failure of empirical treatment with a combination of piperacillin-tazobactam an amikacin. METHODS: 76 febrile episodes with haematological malignancies, neutropenia and resistant fever to the combination piperaciline-tazobactam and amikacin were randomised to receive either teicoplanin or vancomycin. Evaluation of clinical efficacy depended on whether or not apyrexia was obtained after 48 hours, after 7 days or at conclusion of aplasia. In the cost analysis we have considered the following partial costs: drug price, preparation and administration cost, monitoring costs, treatment of adverse events and treatment failure. RESULTS: Primary success of second-line therapy was obtained in 35 cases (46%) with no significant difference between vancomycin (17/38) and teicoplanin arms (18/38). No differences in renal or hepatic toxicity related to the antibiotic therapy was observed. The average cost per patient according to glycopeptide used was $ 450 (180) for the teicoplanin group and $ 473 (347) for the vancomycin group, no significant differences were found in the total cost. CONCLUSION: Our pharmacoeconomic analysis demostrated that teicoplanin and vancomycin can be administered in neutropenic haematology patients with similar efficacy and direct costs.