From a study area in Aust‐Agder, southern Norway, game reports from 1920–1978, supplemented with autumn counts carried out during 1968–1984, were used to determine whether the autumn population size of capercaillie Tetrao urogallus showed no increase, a slight increase or a strong increase compared to the population size the previous year. Based on the mast depression hypothesis, it was predicted that adverse weather conditions should have less influence on the reproduction and thus also on the autumn population size in postmast years of bilberry Vaccinium myrtillus than in other years, because of a higher food quality and therefore also a higher body condition of the birds. In a logistic regression model, only the bilberry index and the June–September temperature of the previous year contributed significantly to explain the status of the capercaillie population. A negative effect of high summer temperatures in the previous year was highly significant when analysing post‐mast years separately, possibly because bilberry plants were less depressed after a high seed crop if summer temperatures and thus primary production were high. Only when years with high bilberry production were analysed separately, did I find effects of weather conditions which could be assumed to have direct impacts on breeding success, such as snow conditions in spring and precipitation in early summer.