OBJECTIVES: Severe sepsis (SS) contributes to a high economic burden among critically ill patients. The objective of the study was to describe SS population and determine associated costs.  METHODS: Data was drawn from a large IHS (Jan 2002 to Dec 2008) and included two groups of adult patients; SSD: SS or septic shock ICD-9 code (995.92 and 785.52); and NSSD: SS criteria by Angus (2001) and Martin (2003). Patients were evaluated at index hospitalization, 30-days (30D), 30-days-6-months (30D-6M), and 6-months-1year (6M-1YR) post-discharge. Comorbidities, antibiotic use, supportive treatment, mortality, and costs were analyzed. Costs were adjusted to 2009 US Dollars using the overall urban CPI; 0.65 (cost-to-charge-ratio) was applied to charge data. Follow-up cost analysis was limited to patients with complete data at the beginning of each evaluation period.  RESULTS: The cohort included 17,256 patients: 3,229 (19%) SSD and 14,027 (81%) NSSD. The groups had similar baseline demographics and comorbidities. SSD had a higher rate of ICU admission (81% vs. 47%). Nearly all received antibiotics. Supportive treatments included vasopressors (66% vs. 19%) and mechanical ventilation (71% vs. 39%). Mean length of stay (LOS) was 21±2d for SSD and 14±14d for NSSD. Hospital mortality was 40% for SSD and 11% for NSSD; overall mortality at 1-year pdschg was 61% and 39%, respectively. Mean cost of hospitalization was $75K±107K, with SSD costing almost double NSSD ($123K±149K vs. $64K±$91K). In all three follow-up periods, NSSD incurred higher costs ($18K±34K vs. $14K±27K for 30D, $54K±102K vs. $42K±81K for 30D-6M, and $34K±67K vs. $27K±56K for 6M-1YR. CONCLUSIONS: SSD patients have higher rates of mortality, both during hospitalization and in the year following, compared to NSSD patients.  The higher mean hospitalization costs for SSD may be due to longer LOS/ more intensive resource use. The lower overall 1-year mortality rate in NSSD may be associated with higher follow-up costs.