Abstract: The paper aims to further improve the haze prediction and control in Linfen City, to avoid and reduce the traffic and health hazards caused by haze weather. According to the observation data of PM_1.0, PM_2.5 and PM₁₀, visibility, relative humidity and temperature from November 2013 to March 2014 recorded by the automatic meteorological station in Linfen City, the authors used data of rainy days and no wind days from December to March during 1954-2014, the mathematical statistics, linear fitting and multivariate regression method, to analyze the relationship between fog haze and meteorological factors in autumn and winter in Linfen City. The results showed that the wind speed, air pressure and temperature would affect the air visibility, wind speed had the biggest influence, wind speed increased per 1 m/s, visibility increased 3.3 km; temperature, relative humidity, wind speed and visibility were negatively correlated to PM_2.5, wind speed and PM_2.5 level had the biggest negative correlation degree, wind speed increased per 1 m/s, PM_2.5 reduced 65.2 μg/m³; no wind days were reducing at a rate of about 0.98 d/10 a, and rainy days were reducing at a rate of about 0.07 d/10 a.