The North China Plain is the political and economic center of China, and it is also the area most seriously affected by groundwater exploitation. Therefore, it is crucial to model and predict groundwater levels in the North China Plain. In this study, using data from groundwater monitoring stations in the North China Plain from 2010 to 2018, wavelet analysis was applied to study the periodicity of shallow groundwater depths in the region, and a spectrum analysis model was established to simulate and predict the groundwater depth in the North China Plain. The results showed a strong correlation between the first significant period and the distance of the station from the Yellow River. The periodic characteristics obtained by the wavelet analysis were substituted into the spectrum analysis model. This improvement significantly increased the accuracy of the modeling results. Moreover, the application of the improved model showed that the groundwater depth in the North China Plain will increase the following year, except for most of Beijing and the coastal areas. Through the study, we understood the dynamic and periodic characteristics of the groundwater level, and proposed a combined method of wavelet analysis and spectrum analysis for groundwater level simulation and prediction, and these results have important guiding for the comprehensive development and protection of groundwater.