Increasing incidence of coccidioidomycosis in California recent years catches our attention. To avoid future widespread contagion and decrease the risk for coccidioidomycosis, statistical models are used to forecast the spread of this epidemic disease. Developed methods for the estimation of Hawkes point process models and recursive point process models facilitate their application for exploring and foretelling the distribution of this type of infection. It is necessary that we can make a preparation in advance when coccidioidomycosis emerges and spreads. We use coccidioidomycosis data from Project Tycho to fit recursive model and Hawkes model with different triggering functions. Serval methods of evaluation are used to value each model’s performance. We also use the first 80% of the data for estimation and the subsequent 20% of the data for evaluation to further check the goodness of models. Three point process models show big similarity, but recursive point process model still does better than others because of its situation-based productivity. We use the recursive model with training and testing data to see how well it can forecast the spread of coccidioidomycosis in California. We demonstrate that the recursive model is shown to fit well to our data and can provide us a reliable prediction.