OBJECTIVES:   While discussions on health technology assessment (HTA) in Japan continue, platforms for real-world population-based studies are lacking.  We attempt to compare two large commercial databases focusing on CHC patient characteristics and treatment patterns. METHODS:  We analyzed the Japan Medical Data Center (JMDC) database containing employer-based health insurance claims and the Medical Data Vision (MDV) database containing administrative data from acute care hospitals. Patients aged 20 years or above and diagnosed with CHC were included in the study. Patients co-infected with hepatitis B and/or human immunodeficiency viruses were excluded. Patient characteristics and treatment patterns for CHC were compared. RESULTS:  3,590 (JMDC) and 29,702 (MDV) confirmed CHC patients were included in the analysis. Patients in JMDC were younger than those in MDV (age 51.6±11.9 and 63.2±13.4) and had fewer comorbidities (Charlson Comorbidity Index 1.3±1.7 and 1.9±2.1). More patients in JMDC were found to have prior CHC treatment (37.2%) than in MDV (24.6%). With an index date of September 2011, about a third of the patients (n=1,271, 35.4%; JMDC and n=9,791, 33.0%; MDV) received treatments for CHC, comprised of 10.2% (n=365; JMDC) and 5.5% (n=1,620; MDV) on antiviral treatments (interferon or direct acting antiviral combinations) and 25.2% (n=906; JMDC) and 27.5% (n=8,171; MDV) on liver protection drugs only. Among patients treated with triple therapy (telapravir/peginterferon/ribavirin), the mean total treatment duration was 19.3 weeks (JMDC) versus 21.9 weeks (MDV). The mean duration of telaprevir therapy was 10.8 weeks (JMDC) and 10.1 weeks (MDV), followed by 11.7 weeks (JMDC) and 14.1 weeks (MDV) of peginterferon/ribavirin therapy. CONCLUSIONS: Both databases found low treatment rates for CHC. Although the findings are consistent, there are differences in database populations and treatment patterns that warrant further research. Using these administrative databases for real-world research may be useful depending on research objectives.