OBJECTIVES: To compare the yearly expenditure per capita of the following high-cost health care programs available in the Brazilian public health care system: severe Rheumatoid Arthritis, renal replacement therapy, Hepatitis, Aids and Oncology. METHODS: A survey in the official health care database (www.datasus.gov.br) was conducted to gather: 1) the amount of resources spent (annual budget), and 2) the number of patients treated in each program in 2007. Those data were then crosschecked with some public information found in the Ministry of Health website (www.saude.gov.br) and from medical societies. For the Aids program an article published by Grangeiro et al in 2006 was used since it resulted from an extensive research and both data on drugs expenditure and the number of patients in 2007 were not available in the official health care database. Based on that article the annual budget and number of patients for the Aids program were updated by the official inflation rate and the official incidence rate, respectively. Then the yearly expenditure per capita for each program was calculated. A limitation of this study is the expenditure per capita calculated for the Oncology program. Since there is no official data regarding the prevalence of oncological diseases in Brazil only new cases (incidence) were considered for the calculation, so the final expenditure per capita is likely to be lower than the one calculated in this study. RESULTS: Severe Rheumatoid Arthritis: R$82,060,720 for 2,098 patients = R$39,120 per patient per year (the highest expenditure per capita); renal replacement therapy: R$1,397,958.956 for 73,605 patients = R$18,993 per patient per year; Hepatitis: R$295,428,000 for 44,598 patients = 6,624 per patient per year;  Aids program: R$1,062,316,834 for 168,417 patients = R$6,308 per patient per year; Oncology: R$1,026,017,544 for 472,000 patients = R$2,174 per patient per year  (the lowest yearly expenditure per capita). CONCLUSIONS: Findings indicate significant differences in the yearly expenditure per capita for each program. Further studies are recommended to extend this analysis and to explore the reasons for those differences.