OBJECTIVES: To estimate cost-effectiveness of adding human papillomavirus (HPV) vaccine to Brazil National Immunization Program, compared with the current strategy for control of cervical cancer, the screening program based on conventional Papanicolaou citology. METHODS: We used a static decision tree analysis model, CERVIVAC, software developed by PAHO ProVac Initiative. The model compared two strategies: 1) a cohort of girls aged 10 years to represent the epidemiological situation, current screening, and care of precursor lesions and cancer (without vaccination) with 2) a cohort of girls aged 10 years who were applied 3 doses of HPV vaccine, maintaining the conditions of screening and care of precursor lesions and cancer, as they are currently. National parameters for the epidemiology of cervical cancer, screening expenditures and National Cervical Cancer Control Program were estimated in depth. The estimates were based on the health information systems of the Public Health System, SUS (SIH, SIA, APAC, SIAB, SIGTAP and SISCOLO), national survey (PNAD 2008) and relevant national literature. RESULTS: The vaccine introduction would reduce the burden of disease. The model estimated 261 deaths avoided and 7595 DALYs averted, in each successive cohort to be vaccinated. The ICER per DALY averted R$ 18,121 would be considered cost-effective, according to the parameters adopted by WHO. In univariate sensitivity analysis, only a lower discount rate and a high incidence and mortality would make the introduction of the vaccine very cost-effective. CONCLUSIONS: The vaccine introduction should bring benefits for the control of cervical cancer, but it will require large investments by the NIP. It should be guaranteed the right conditions for the immunization program sustainability and equity in a population perspective as well as to improve the screening program.