Modeling price fluctuations in financial markets is very important. We try to model price fluctuations in Tehran stock exchange using heterogeneous agents’ model.  We used agent-based computational approach. In this model, there are two kinds of agents, some agents have extrapolating expectations (chartists) and others have stabilizing or mean-reverting expectations (fundamentalists). The dynamics of shares of these two types of agents make price fluctuations. For determining the relative effectiveness of agents expectations, Diechi and Westerhoff (2012) method, is used.  For this purpose, weekly data of Tehran Stock Exchange price index (TEPIX) from 1997 to 2014 were used. Modeling results show that the relative sensitivity of buyers with different expectations, and their relative impact to aggregate demand, have significant and important role in the price dynamics of Tehran stock market. We also show that the relative impact of chartists to price fluctuations very important and over the past two decades, the share of them from aggregate demand have been more than 80 percent.