Background Yellow fever is endemic in Africa and the Americas, occurring in urban or sylvatic environments. The infection presents varying symptoms, with high case-fatality among severe cases. In 2016, Brazil had sylvatic yellow fever outbreaks with more than 11 thousand cases, predominantly affecting the country’s Southeast region. The state of Minas Gerais accounted for 30% of cases, even after the vaccine had been included in the immunization calendar for at least 30 years. Methodology and principal findings We applied parameters described in the literature from yellow fever disease into a compartmental model of vector-borne diseases, using namely generation time intervals, vital host and vector parameters, and force of infection, using macroregions as the spatial unit and epidemiological weeks as the time interval. The model permits obtaining the reproduction number, which we analyzed from reported cases of yellow fever from 2016 to 2018 in residents of the state of Minas Gerais, Brazil. Minas Gerais recorded two outbreak periods, starting in EW 51/2016 and EW 51/2017. Of all the reported cases (3,304), 57% were men 30 to 59 years of age. Approximately 27% of cases (905) were confirmed, and 22% (202) of these individuals died. The estimated effective reproduction number varied from 2.7 (95% CI: 2.0–3.6) to 7.2 (95% CI: 4.4–10.9], found in the Oeste and Nordeste regions, respectively. Vaccination coverage in children under one year of age showed heterogeneity among the municipalities comprising the macroregions. Conclusion The outbreaks in multiple parts of the state and the estimated R_e values raise concern since the state population was partially vaccinated. Heterogeneity in vaccination coverage may have been associated with the occurrence of outbreaks in the first period, while the subsequent intense vaccination campaign may have determined lower R_e values in the second period.