OBJECTIVE: The purpose of this study was to determine the cost effectiveness of Erythropoietin against transfusion in HIV patients treated with Zidovudine from the societal perspective. METHOD: Cost-effectiveness was measured based on probability of response to dose. The cost of drug with and without the patient drug expenditure cap under the Manufacturers’ Patient Assistance Program was considered. Measured were incremental cost per unit of blood spared and incremental cost per QALY. Sensitivity analysis was performed on all variables in the base case. RESULTS: Without the drug expenditure cap, the incremental cost per unit of blood spared was $577.93 ($160.34 with the cap). The incremental cost per QALY without the cap was $118,620.13 and $32,909.71 with the cap. The cost per unit of blood with the cap was 28% below the cost of a unit of blood (PRBC). The break-even cost of the drug was 77% below the current cost. A plot of the cap against the cost per unit of blood spared yielded a straight line. The cost per unit of blood spared became zero at $4,363.25 and became equal to the cost of a unit of blood at $10,145. CONCLUSION: With the cost cap, erythropoietin can be justified by incremental cost per QALY and on its blood sparing and cost saving effects. The drug is most cost effective when it obviates the need for transfusion entirely while it cannot be justified in patients who would otherwise require 3 units of blood or less per month. Variations in the cost of transfused blood units impact cost-effectiveness but remain within acceptable limits for typically cited values.