In the stock exchange, stock prices prediction is an important issue for traders in the capital market to assist make decisions in the transaction. One method that has artificial intelligence techniques can be applied for forecasting or prediction is fuzzy method . The approach used in forecasting stock price research is fuzzy Mamdani. With the toolbox of Matlab, the process of forecasting with fuzzy Mamdani can be easier. The expectation of using fuzzy Mamdani models is to improve forecasting accuracy results in response to uncertainty. Afterwards this model can predict the stock prices one day within a certain time period in order to obtain the best prediction results. Forecasting results using fuzzy on the Forbes five best companies in 2012 resulted has Mean Absolute Percentage Error is less than 10 % for a span of one week and one month, and less than 20% for the span of one year and three years.