OBJECTIVES: A universally accepted staging system for hepatocellular carcinoma (HCC) has not been adopted. The goal of this work is to assess survival and treatment patterns of HCC patients stratified by the National Comprehensive Cancer Network (NCCN) staging-guided treatment categories. METHODS: An HCC retrospective research registry was developed from linking records in the University of Utah Enterprise Data Warehouse, the Utah Population Database, and the Huntsman Cancer Institute Tumor Registry. Patient records were included if age >= 18, at least one HCC ICD-9 code, documented hepatocellular carcinoma or cholangiocarcinoma in the tumor registry, and the presence of informative staging data around the time of diagnosis. Patients were stratified by the NCCN groupings by chart review around the time of diagnosis as either: potentially resectable or operable (RESECT), transplantable (TRANSP), unresectable (UNRESCT), inoperable due to performance status (INOPER), or having metastatic (METAST) disease. Treatment patterns and survival were evaluated. RESULTS: A total of 221 patients (161 males, 72.9%) with HCC were identified. At the time of diagnosis 12.7% subjects (n=28) were considered RESCT; 14.9% (n=33) TRANSP; 34.8% (n=77) UNRESCT; 18.1% (n=40) INOPER, and 17% (n=38) METAST.  Staging information was not available for 5 subjects (2.3%). Kaplan-Meier analysis demonstrated statistically significant differences in all groups except between RESCT and TRANSP patients and between UNRESCT and INOPER patients. The median days of survival were 594, 562, 247, 167 and 44 for RESCT, TRANSP, UNRESCT, INOPER, and METAST patients respectively. RESECT patients most frequently received early resection (61%) and TRANSP patients had the highest utilization of liver transplant (33%). Utilization of any treatment was low in the METAST (31.6%) and INOPER (60%) groups. CONCLUSIONS: Treatment patterns were distinct across the NCCN groupings and aligned with NCCN treatment guidelines. Overall the NCCN groupings predict overall survival.