The introduction of the temporal component into poverty analysis takes us to the study of the poverty dynamic, which focuses on investigating events of entering, remaining in and getting out of deprivation. As an effort to consider the diverse aspects unrelated to income in the study of vulnerability to poverty, the monetary approach is put aside and the multidimensional approach is used to elaborate the well-being index. The Multidimensional Poverty Index (MPI) is constructed with data from Brazil’s 2000 and 2010 Demographic Census. First, this article intends to quantify and describe the results of the MPI in Brazilian territory, using municipalities as unit of analysis. Second, it intends to analyze the factors associated to the poverty dynamics, which can be understood through the transition probability from a non-vulnerable state to a vulnerable state, in a distribution of the MPI in deciles. Finally, it intends to identify the municipalities that went through this transition in a regional level during the intercensal period. Results show that, despite the improvement observed in its whole national territory, Brazil’s North and Northeast remain with deeper deprivations while the Southern and Southeastern regions present the lowest incidence of multidimensional poverty. Dropout Rate of Grades 3–4 of High School, HDI-M Education, HDI-M Longevity, Logarithm of the Population, Gross Value Added per capita of Industry, Service Concentration Index and Entrepreneur Rate contribute positively to the reduction of municipal vulnerability to poverty. Compared to Northern municipalities, Southeastern Brazil has reduced more significantly the likelihood of position loss and has had a greater effect of increasing the probability of its upward movement.