Abstract—The exchange rate of IDR of USD can be the indicator to analyze Indonesian economy. The exchange rate which is an important factor has big effect in Indonesia economy overall. The government gives attention to the movement of exchange rate. So, it needs a mathematics model to forecast the exchange rate of IDR of USD. The model that used is fuzzy wavelet model. Fuzzy wavelet model is the combination of fuzzy Mamdani model and Discrete Wavelet Transforms (DWT). DWT is used to transform a function with the time domain into the frequency domain. This model is very effective to identify the case, have high accurate result, and have simple structure. In this paper, data of exchange rate that are used are weekly data taken from January 1, 2012 until November 9, 2014. The result of forecasting gives a level of accuracy of MAPE of 3.45% for training data and 1.74% for testing the data.