The main purpose of this research was whether compensation system in education organization of Zandjan province can distinguish between employees based on their personal characteristics. To meet the purpose, of 12950 monthly paid teachers in different age, academic paper, organizational positions and length of service, 2554 of employees were selected based on cluster random sampling in 5 clusters that each cluster represents one area. To analyze the research questions, descriptive and inferential statistics like path analysis and discriminant analysis using Spss 11.0 were served. The relationship between outcome variable, monthly salary, and job category is 0.73 and between monthly salary and length of service is 0.815. Both of them were significant (P<0.01), showing that there is a positive and direct relationship between variables. Finally the causal model encompassing personal variables influencing the salary is drawn.