In Iran, regional unemployment rates appear to vary widely. In the year 2006, unemployment rates in Sistan and Balochestan, and Yazd provinces were 31.5 and 7.87 percent, respectively. This disparity is widely in county level. For exampl, unemployment rate at Saravan county in the Sistan and Balochestan province was 59.87 percent, while this rate was 0.15 percent at Abomosa county in the Hormozgan province. Because of regional differences in unemployment rate, geographical analysis is important.In this paper, by using spatial econometrics models which is applied in spatial autocorrelation and Iranian census data in the year 2006, we analyze the regional differences unemployment rates among the 336 cities.On the basis of findings, there were two phenomena: there exist a positive spatial dependency or autocorrelation and geographical heterogeneity in Iranian labor market at the provincial level. Provinces marked by high unemployment rate, as well as those characterized by low unemployment rate, tended to be spatially clustered, demonstrating the presence of ‘spatial persistency’.