Increasing the quality of human resource via education investment is a sine qua non for welfare promotion as well as for poverty solution. Education, inflation, income per capita, consumption, regional gross domestic product, and economy growth are a number of variables that directly and indirectly affect the poverty. The research uses time series data of the last ten years (1999-2009) obtained from Central Agency on Statistics (BPS). The aim of this research is to identify the causal relationship of the influential variables mentioned. The path analysis is used to analyze the data based on the five models. The simultaneous testing of table summary shows that the Regression coefficient value of each model is significant, where F value is smaller than alpha 0,05. The similar conclusion is also shown in partial testing between the independent variables and dependent variables.