In recent years, anthropogenic emissions in the Fenwei Plain (FWP) have decreased; however, haze pollution remains a serious issue. This study explored the possible reasons for this enduring problem in terms of climate and meteorology. Firstly, the contribution of climate and meteorology to haze pollution in the FWP was quantified using a best fit model and differences in key meteorological parameters were analysed over several time periods. Key climate factors were identified using a relative importance test and correlation analysis, and the adjusted optimal subset model (AOSM) was used to predict the number of winter haze days for the entire winter and for individual winter months, respectively. Results showed that the average minimum contribution (CONave) of climate/meteorology to winter haze pollution was 24.3% from 1984 to 2016 and the modulating role of climate and meteorological conditions increased significantly after 2010 (CONave = 55.8%). This was attributed to a significant decrease (increase) in sea level pressure and surface wind speed (500 hPa geopotential height and surface temperature) in the FWP. The explained variance (R2) and mean absolute error (MAE) of the models for the entire winter period were 88% and 3.54 days, respectively. For predictions in November, December, January, and February, the R2 was above 71% and the MAE was below 1.83 days. Furthermore, independent predictions for 2018 showed that the bias in the monthly models was lower by 5 days compared to the entire winter model. The results of a recycling independent test indicated that all the models evaluated had excellent stability, proving that the climate factors chosen by the AOSM do affect haze pollution in the FWP.