Abstract: According to the synchronization of extremely dense fog (EDF) at individual stations, Anhui province can be divided into five regions with different statistical characteristics of fog. To understand the evolution and causes of regional extremely dense fog (REDF) in Anhui province, the criteria to identify REDF days in Anhui province are determined first based on the 08:00 BT visibility data and daily weather phenomenon records collected at 68 national meteorological stations that have complete data in Anhui province from 1980 to 2019. Time series of REDF days in each region over the 40-year period are obtained, and the interannual and interdecadal variation trends of REDF are analyzed. Monthly and daily variations and duration of REDF in different regions are then analyzed using hourly data collected at 77 national meteorological stations from 2016 to 2019. Finally, the causes of interannual variation of REDF in winter are investigated. Results are as follows: (1) From 1980 to 2019, annual REDF in the three regions along and to the north of the Huaihe river first increases and then decreases with the turning point in 2006/2007. The obvious annual REDF increase from 1980 to 2007 may be attributed to the increase in aerosol particles during the same period. Decadal average number of REDF in all the regions is the highest in the 1990s and/or 2000s and the lowest in the 2010s. The number of REDF in each region varies greatly from year to year, with a minimum of 0—1 d and a maximum of more than 10 d. (2) From 2016 to 2019, the number of annual average REDF days in each region is 14—17 d, and these days are mainly concentrated from mid-autumn to mid-spring. Those REDF days with fog lasting for only 1 h account for the largest proportion of total sample days, followed by those with fog lasting for 3 h. The two regions to the north of the Huaihe river have the largest number of annual REDF days and also the largest percentage of REDF days with fog lasting for at least 3 h. (3) The number of REDF in winter in regions to the north of the Huaihe river is significantly positively correlated with the number of precipitation days, seasonal precipitation, relative humidity and air temperature at 08:00 BT. But the correlation with wind speed and the number of light wind days (daily average wind speed less than 2 m/s) is not significant. The number of winter REDF days in the region along the Yangtze river is mainly determined by ground level wind speed. However, there is no significant correlation between the number of winter REDF days and any ground level meteorological factors in the region to the south of the Yangtze river. (4) Taking January as an example, REDF days are positively correlated with zonal circulation index in each region. The REDF days in the three regions, including one region along the Huaihe river and two regions to the north of the Huaihe river, are positively correlated with the location of the East Asian trough but not with its intensity, suggesting that zonal circulation and eastward location of the East Asian trough contribute to the formation of REDF therein. Further comparative analysis of the circulations in January between more REDF days and less REDF days in consecutive years in the region to the north of the Huaihe river shows that, when the 1030 hPa isoline of sea level pressure to the north of 40°N is located to the east (e.g. to the east of 120°E) in January, more REDF days could occur with stronger easterly winds and higher relative humidity at ground level.