The study aims to explorethe time series dynamic relationship between meteorological conditions and the yield of chili peppers in the Qiannan region,and lay a solid theoretical foundation for production planning, climate adaptability analysis, and future yield prediction of the chili pepper industry in this area. Four methods, namely the single exponential smoothing method, the quadratic exponential smoothing method, the 3-year moving mean method, and the HP filtering method, were adopted to separate the unit yield data of peppers in Qiannan region from 2013 to 2022 into the trend yield and meteorological yield of peppers. Through correlation analysis, the main meteorological factors affecting yield of pepper were screened out. Subsequently, a stepwise regression model was used to construct a prediction model for the meteorological yield of peppers. According to research findings, (1) four different methods for separating yield all effectivelyreflected the consistency characteristics of the meteorological yield sequence and the changes in meteorological disasters in Qiannan. Among them, the trend yield separated by the single exponential smoothing method could efficiently capture the more intense fluctuation parts in the yield; (2) in the meteorological yield prediction model of peppers, except for Duyun and Guiding, the R² of the meteorological yield prediction model of peppers in the other counties ranged from 0.728 (Longli) to 0.998 (Fuquan), and all passed the 99% confidence test. Except for Pingtang, Sandu, Fuquan and Duyun, the prediction models based on meteorological factors constructed for the remaining counties performed well in simulating the yield of peppers, and the model accuracy and prediction accuracy were relatively high, which can be applied to the estimation of pepper yields.