Economic evaluation of medical technologies in Japan is challenging because of a lack of data sources. OBJECTIVE: This paper reviews the data sources available for conducting pharmacoeconomics research in Japan, and discusses health policy reforms that may increase the demand for pharmacoeconomics studies in the future. METHODS: A literature review of pharmacoeconomic studies was conducted in both English (MEDLINE OHE-HEED) and Japanese publications (JAPICDOC) to identify the typical data sources used in these studies. Reflecting on first hand experience conducting outcomes research in Japan, we summarized (percent of studies utilizing the different data source types) and commented on the data sources. RESULTS: A total of 42 published studies were found in the Japanese and English literature from 1980-2001. Typical sources identified for resource utilization data included: clinical trials (62%), epidemiology/case studies (43%), patient chart reviews (52%), expert panels (12%). Sources for clinical efficacy or disease state transition probabilities included: clinical trials (60%), epidemiology/case studies (36%), patient chart reviews (29%), expert panels (12%). The cost data sources identified were: 1) national uniform reimbursement fees (98%) - Japanese Ministry of Health, Labor and Welfare (MHLW), 2) micro-cost studies (2%). The MHLW uniform fees represent costs from the health insurance plan viewpoint, and therefore are relevant in drug price negotiations with the MHLW. Demand for pharmacoeconomic data may increase because of: 1) upcoming MHLW pharmaceutical pricing and/or health insurance system reforms; 2) new regulatory committees e.g. NICE. CONCLUSION: Access to data sources for pharmacoeconomic research is generally known to be limited in Japan, however several published studies effectively overcame such limitations. Compared to other countries, however, detailed population-based data are still relatively unavailable. Upcoming health care reforms may increase the demand for pharmacoeconomic data in Japan, thus, increasing the demand for better sources of data typically needed for these studies.