Given the scarcity of water resources and the importance of their optimal management, accurate prediction of groundwater level fluctuations is essential. Intelligent models such as time series, wavelet analysis, artificial neural networks, and support vector machines can help in the sustainable use of groundwater. Objective: The objective of this study is to use a combined wavelet-fuzzy neural model to predict groundwater level in the Birjand Plain and compare its results with wavelet and fuzzy neural network models. Methods: In this study, rainfall, evaporation, temperature, humidity, and groundwater level data from 16 piezometers for 18 statistical years (monthly) were used. The combined wavelet-fuzzy neural model, wavelet model, and fuzzy neural network were used to predict groundwater level, and their results were compared with each other