Based on the ground-timing observation data in the Beijing—Tianjin—Hebei (BTH) region, the surface Eurasian snowpack data, the ECMWF and the NCEP/NCAR reanalysis datasets,and the Hadley Centre's Arctic sea ice concentration data from 1980 to 2017, the present study investigates the physical linkage between the interannual variations of haze frequency over the BTH region in October—November and the concurrent Eurasian snowpack. Furthermore, the connected potential mechanisms are also discussed via the climatic statistical diagnosis and numerical sensitive experiments. Results show that the interannual variability of the October—November haze frequency over the BTH region shows a significant positive correlation with the simultaneous snow depth as well as the snow-cover extent over the eastern Europe—Siberian plain (R_Eufor short; 50°—60°N, 40°—80°E). Those localized positive snowpack anomalies are closely tied to the anomalous atmospheric diabatic cooling areas to its northwest side, i.e. Norwegian Sea—Barents Sea sector and the northern and eastern European regions, which can in turn excite a large-scale zonal quasi-barotropic Rossby wave train from the area around the key area R_Eu to the Northeast Asia, modulating the climate variability in the concurrent Northeast Asian anticyclonic anomaly, a key circulation system leading to a higher number of the simultaneous haze days in the BTH region. Under such circumstances, the lower troposphere over the BTH region is dominated by anomalous southerly winds. The in situ stable stratification can be easily set up, and the boundary layer height is lower than normal, along with the weakened surface wind velocity and enhanced surface relative humidity. The above environmental conditions are conducive to the occurrence and development of localized haze weather, thus resulting in more frequent haze events.As a precursory signals, the preceding decrease in the September Chukchi—West Beaufort Sea ice may lead to a higher number of October—November haze days over the BTH region,and vice versa.