Introduction: Poverty, education and economic growth are three macro-economic variables that sometimes are closely related together and sometimes their relation is complicated. Accordingly, the present study has been tried to identify and modeling the relationship between these three variables.

Method: To achieve this goal, firstly using by Granger causality relationship between these three variables has been identified and then for the period 1363-1389 is modeled in the framework of Auto Regressive Distributed Lag Method (ARDL).

Findings: Granger causality test results showed the direction of causality is from economic growth and education to poverty. Modeling the results of Granger causality test in framework of the ARDL model also shows that poverty, education and economic growth are co-integrate. In other words, long-term relationship exists between them. Also, the results of this model show that the education with one lag in short and long term has a negative and significant effect on poverty indicator in Iran. Also, the relationship between economic growth and poverty is reverse in the short and long term. Finally, the result of error correction model (ECM) showed the speed of adjustment toward long-run equilibrium in the poverty model slowly being carried out.

Discussion: with increasing economic growth and education, measuring of poverty in the society will decrease in the short and long term. Also, if a shock is inserted into the model, two periods are needed to model adjust itself, and again to get the initial equilibrium.