Abstract: In order to improve the fog forecast accuracy in Ma’anshan, the authors analyzed the climatic characteristics of heavy fog in Ma’anshan from 1961 to 2010 and established a fog prediction method based on the method of ingredients. The results showed that: the fog days of Ma’anshan focused on November to January of the following year, the fog days in winter were more than that in summer; the humid air (relative humidity was 90% to 100%), the weak wind (wind speed was equal or less than 3 m/s) and suitable temperature (equal or less than 20℃) were all beneficial to the formation of fog; whether there was an inversion layer in 925 hPa and 1000 hPa was beneficial to the thermosphere and was extremely important for the formation and maintenance of fog; the ground situation of heavy fog in Ma’anshan was divided into four categories: weak high pressure type, the back type of the high-pressure sea-going, cold front warm area type, ground reverse grooved type. The authors chose the water vapor condition, cooling condition, stratification condition and wind conditions as the basic ingredients of fog forecast and established the prediction method of Ma’anshan. Using the data of EC fine grid (September 2011-December 2013) to analyze the failure report and empty report for this validation method: TS score was 42.86%, non report rate was 6.9% and the empty report rate was 55.74%.