SummaryDaily winter hive temperatures were recorded in multiple-packs of honeybee colonies (Apis mellifera) equipped with different types of hive entrance. Sixteen hives, each of two Langstroth boxes, were divided into 4 groups; these were then packed with insulation and tar-paper for wintering. In each group, two hives faced N and two S and the same treatment was used within each group. Treatments were: (1) bottom and top entrances (1×5 cm each); (2) fully open bottom entrance; (3) bottom and upper-side entrances (1×5 cm each); (4) bottom entrance (1×5 cm) and a 2·5-cm diameter auger-hole in the middle of the second chamber. During the 1979–1980 and 1980–1981 winters the temperature was recorded daily at 08.00 h using thermistors at 54 points in the SW hive in each of the groups. The ambient temperature fell as low as−38·8°C but the lowest hive temperature did not go below a range from−1°to−14°.All colonies in treatments 1 and 4 maintained a temperature of 30°or higher in the winter cluster throughout both winters. All colonies occupied the upper hive body; colonies in treatment 1 occupied the area close to the top entrance and in treatment 4 the area close to the auger-hole. The temperature of the cluster in colonies in treatments 2 and 3 fell to 28°or 29°on 2 days in December 1980. Colonies with the side top entrance (treatment 3) appeared to be unable to move consistently to that entrance. Colonies in treatment 2 appeared to move slowly to the upper hive body; whilst ambient temperatures remained relatively high they developed rapidly, but with a consequent drop in temperature they lost more cluster size than other colonies. The large quantity of stored food and absence or scarcity of empty cells appeared not to hinder cluster movement.