OBJECTIVES: To compare the availability and quality of data needed for pharmacoeconomic analysis in Brazil versus the US and EU, using diabetes and hypertension as case studies. METHODS: An extensive search was undertaken in Brazil for sources of epidemiologic, economic, outcomes, and life expectancy data needed for pharmacoeonomic analysis in the areas of diabetes and hypertension. Potential sources included the published literature, holdings of national health authorities and government statistical agencies, public and commercial insurers, pharmacy benefit management companies, and disease management firms. Identified data sources were evaluated according to five criteria: content; coverage; quality; reliability; and timeliness. Sources identified in Brazil were then compared with corresponding sources commonly used in the US and EU. RESULTS: Many of the data elements needed for pharmacoeconomic analysis were identifiable in Brazil and comparable to US and EU sources on the five evaluation criteria. These included the costs of prescription drugs and health-care services used in the treatment of diabetes and hypertension as well as age-specific mortality rates and life expectancies, which are freely available from the government data system (DATASUS) and available for purchase from selected commercial database vendors. Notable gaps in requisite data in Brazil included large-scale longitudinal epidemiologic studies analogous to the United Kingdom Prospective Diabetes Study (UKPDS) and the US Framingham Heart Study, needed to estimate longitudinal risks of diabetes- or hypertension-related cardiovascular events. In addition, country-specific utility data of relevance to diabetes and hypertension were not found. CONCLUSION: Brazilian data sources are available and of sufficiently high quality to be used in pharmacoeconomic analyses in diabetes or hypertension. However, in the absence of Brazilian longitudinal epidemiologic data and disease-specific utilities, scientifically rigorous analyses for the Brazilian setting likely would require local adaptation of models and data from other countries in North America or Europe.