Intracerebral hemorrhage (ICH) is a devastating form of stroke, resulting in mortality and disability. A recent Phase IIb clinical trial has shown that recombinant factor VIIa (rFVIIa, NovoSeven) significantly reduces mortality in ICH patients and improves functional outcome.  OBJECTIVES:  To estimate short-term cost-effectiveness of rFVIIa compared to standard care in treating ICH from a US hospital (inpatient) perspective.  METHODS:  A decision-analytic model was developed to estimate the cost-effectiveness of rFVIIa 40mcg/kg, 80mcg/kg, and 160mcg/kg compared to standard care in treating ICH from a hospital perspective. Costs and outcomes were estimated for a patient's initial hospitalization.  Mortality, disability, and initial hospital length of stay (LOS) was obtained from the Phase IIb clinical trial. Direct medical costs for initial hospitalization following ICH were assumed to include all costs associated with inpatient care estimated from an analysis of Medicare claims data. rFVIIa costs were based on average sales price.  Costs were in 2005 US dollars. Sensitivity analyses were conducted to assess robustness.  RESULTS:  Treatment with rFVIIa 40mcg/kg and 160mcg/kg resulted in additional costs of $2283 and $6700 respectively compared to standard care, which includes the cost of rFVIIa, after factoring in relevant inpatient costs.  Treatment with rFVIIa 80mcg/kg was associated with a reduction in expected medical costs (-$333).  Given the clinical trial results, the incremental cost-effectiveness ratio (ICER) per survivor for rFVIIa 40mcg/kg, 80mcg/kg, and 160mcg/kg were $19,726, -$3,178 (cost-savings), and $68,723 respectively.  In addition, average costs per survivor for rFVIIa 40mcg/kg, 80mcg/kg, 160mcg/kg, and standard care groups were $44,102, $41,475, $50,582 and $48,085, respectively. Results were robust to changes in model parameters.  CONCLUSIONS: Treating ICH with rFVIIa 80mcg/kg is not only cost-effective but also cost-saving to the hospital in the short-term compared to standard care.  Cost-effectiveness results were driven by treatment efficacy (mortality), LOS (and corresponding inpatient costs), and rFVIIa costs.