This research aims to analyze nine models of poverty determinants in seven regions in Indonesia. The formation of the model is based on model 1, namely the poverty trap, economic growth, DAK, and spending on government functions which consists of spending on education, health, economic functions, social protection, as well as housing and public facilities. The next model was formed by combining the economic growth variables and DAK with GDP per capita and real GDP, as well as DAK with DAK for 1 year and DAK for the previous 2 years, while the poverty trap and government function spending remained the same and were not combined. In this way, nine poverty determinant models were obtained which were applied to seven regions in Indonesia, namely Sumatra, Java and Bali, Kalimantan, Sulawesi, Nusa Tenggara, Maluku and Papua. This research uses secondary data from BPK and BPS. The data used is panel data from 509 districts/cities (cross section) in 2014-2020 (time series). The method used is panel data regression analysis. This research produces a model of the determinants of poverty in each region. The determinant model for the Sumatra region is model 1, Java and Bali are models 6 and 9, Kalimantan is models 2 and 3, Sulawesi is models 4 and 7, Maluku is models 1 and 2, and Nusa Tenggara and Papua are models 2 and 3. The poverty trap variable is significantly has a positive direction in all regions, which means that the worse the poverty level was in the past year, the current poverty level will increase. Therefore, to overcome poverty levels in the coming year, the significant variables produced in this research model can be used as a reference.