By using thunderstorm gales observations in summer in the period of 2009—2015 and NCEP reanalysis data, this study divides the thunderstorm gales environment into dry and wet environment. The results show that the thunderstorm gales days in wet environment account for about 86% of all thunderstorm gales days. Based on physical quantity parameters and Logistic regression method, potential forecast models for regional thunderstorm gales in dry and wet environment during summer over Yangze-Huaihe River Basin are developed. The most significant predictors for thunderstorm gales in wet environment over southwest, southeast and north regions are CS, K and SI, respectively. The most significant predictor for thunderstorm gales in dry environment is TT. Compared to WINDEX, the new wind index (GUSTEX) which considers both thermodynamical effect and high-level horizontal momentum information shows better predictive significance for thunderstorm gales in dry and wet environment over Yangze-Huaihe River Basin. The probability thresholds are determined through historical sample return. The independent samples of 2016 are used to verify the model and the Logistic model is proved to have a good performance.