OBJECTIVES: Two major clinical trials examined the efficacy of caspofungin (CAS) and voriconazole (VORI) for empiric therapy of febrile neutropaenia (FN). We investigated the cost-effectiveness of empiric CAS vs. VORI in FN from the Turkish perspective. METHODS: The downstream consequences of CAS or VORI were captured through decision tree analysis. Outcome measures included success, breakthrough fungal infection, persistent base-line fungal infection, persistent fever, premature discontinuation and death. Probability data were extracted from the major studies. An expert panel estimated healthcare resource consumption and alternative treatment after initial failure with either agent. Cost was based on 2012 data using Turkish Lira (TL). Deterministic and probabilistic sensitivity analyses were performed. RESULTS: Compared to VORI, CAS was dominant by TL2,533, TL29,256 and TL2,536 per patient treated, successfully treated and patient survival, respectively (approx. USD1,414, 16,328 and 1,415). CAS had a higher likelihood of success and lower mortality than VOR (34.17% vs. 26.02% and 7.37% vs. 7.95%, respectively). Increasing the list cost or length of stay (LOS) for CAS by >35% or 1.3 days, respectively, changes the study outcomes. A decrease of list cost or LOS for VOR by >32% or 1.2 days resulted in it being favorable. Removing fever resolution as part of the composite outcome afforded a contracted difference (CAS preferred by TL298 and 299 per patient treated and surviving with VORI preferred by TL488 per patient successfully treated). Monte Carlo simulation of 10,000 subjects, with variability imputed on the outcome probabilities taken from the literature, LOS and hospitalisation costs, resulted in a 78.8% chance of favoring CAS. CONCLUSIONS: There is a high likelihood of CAS being cost-effective compared to VORI in the treatment of FN in Turkey. Sensitivity analyses highlighted a robust advantage towards CAS. The model is moderately sensitive to changes in LOS or cost of each agent.