On the basis of many years of research (1972-2021), the peculiarities of counting waterfowl at the largest “cold” wintering in Eastern Siberia in the source and upper reaches of the Angara river are considered (South Baikal). Long-term practice of work shows that for a more accurate determination of the number of wintering birds, it is necessary to use special corrections to the materials obtained on the basis of an absolute registration of birds. One of the main corrections is the proportion of birds under water at the time of counting (correction for diving). The fact is that birds constantly dive while foraging and at any the time some part of them is not available for accounting. In the conditions of a short winter day (about 8 hours), he almost entirely goes in search of food. The magnitude of this correction depends on many environmental factors, but by now its relationship with the surface air temperature and the length of daylight hours has been proven. It is largely by these two factors determined. The multiple correlation coefficient between these features is R1.23=0.46 and highly reliable, P < 0.05. From the time of the formation of “cold” wintering (midJanuary) to spring (mid-March), the value of the correction decreases – from 45.0 % to 35.0 %. It is constant fairly, despite the current pronounced trend towards climate warming. Only at the beginning of the formation of wintering (about 10 days), associated with a period of particularly «severe» weather conditions, can its value exceed 50.0 %. However, at this time, it is very difficult and irrational to conduct surveys due to extremely unstable ice conditions. As the length of daylight increases (up to 12 hours in mid-March) and the temperature of the surface air layer rises, the activity of feeding on birds decreases. The size of the correction is also decreasing – to 35.0 %, and in April to 28.9 %. This, to extent a large, is associated with a noticeable increase in the comfort of wintering conditions for birds in the «cold» wintering. On average, over the entire observation period (JanuaryApril), the magnitude of the correction is 37.8 %. However, this value cannot in bird counts be used. It is necessary to apply the corrections determined for each month of the «cold» wintering. At the same time, this approach is highly formalized. Long-term observations during field work and analysis of the dynamics of weather factors, first of all, surface air temperature, show that periods of sharply differing weather conditions over a long study period during one season are well identified on the graphs. Therefore, the boundaries of transitions from more severe weather conditions to more comfortable ones can be easily distinguished. Consequently, it becomes possible to shift the monthly boundaries, within which certain corrections for diving are in effect, for specific periods of more severe or more favorable weather. Of course, in this case, without fail, it is necessary to take into account the general tendency towards a decrease in the correction for diving as the winter conditions become more comfortable during the winter. The use of the correction for bird diving significantly clarifies the materials of accounting works and, along with other amendments, is a necessary element of research to determine the number of wintering birds and the organization of long-term monitoring of the state of this “cold” wintering.