The objective is to assess the cost-effectiveness of pegfilgrastim for the prevention of hospitalization due to febrile neutropenia (FN) in patients with epithelial ovarian carcinoma (EOC) receiving taxane/platinum-based chemotherapy. A decision analysis model evaluated a hypothetical cohort of 10,000 patients receiving six cycles of taxane/platinum-based chemotherapy for EOC. Three strategies were analyzed for the prevention of hospitalization due to FN: 1) dose modifications and delays after a hospitalization for FN without the use of granulocyte-colony stimulating factors (G-CSF) (NO G-CSF); 2) all patients receive G-CSF with each chemotherapy cycle (1 degrees PROPHYLAXIS); 3) patients receive G-CSF for all subsequent chemotherapy cycles after a hospitalization for FN (2 degrees PROPHYLAXIS). The model was applied to two patient populations: 1) an average-risk population (FN hospitalization rate = 5%); 2) a high-risk population (FN hospitalization rate = 16%). Using baseline assumptions in an average-risk population, NO G-CSF was the least expensive strategy with a cost of $68 million and resulted in 2,860 hospitalizations for FN. 2 degrees PROPHYLAXIS resulted in 141 fewer hospitalizations than NO G-CSF at a cost of $76,288 per hospitalization prevented. 1 degrees PROPHYLAXIS was the most effective and resulted in 1,689 fewer hospitalizations for FN compared to NO G-CSF at a cost of $47,343 per hospitalization prevented. When this model is applied to a high-risk patient population, 1 degrees PROPHYLAXIS is more effective and less expensive than both NO G-CSF and 2 degrees PROPHYLAXIS. We conclude that in average-risk patients receiving chemotherapy for EOC the use of pegfilgrastim is effective at reducing hospitalizations due to FN, but at a significant cost. However, in high-risk patients, primary prophylaxis is the only cost-effective strategy and should be strongly considered.