This paper assessed the potential exposure of bees (Apis mellifera L.) to pesticides during maize (Zea mays L.) sowing with pneumatic drills. Data were derived from tests carried out in field tests, comparing two configurations of a pneumatic precision drill: conventional drill; drill with air deflectors. In addition, static tests simulating the sowing under controlled conditions, were performed on the drill equipped with an innovative system developed at CRA-ING. During the field tests, the concentrations in the air of the active ingredients of four insecticides used in maize seed dressing (imidacloprid, clothianidin, thiamethoxam and fipronil) were recorded. The concentrations of active ingredients in the air were used for assessing the quantities of active ingredients that a bee might intercept as it flies in a sort of virtual tunnel, the dimensions of which were dependent upon the bee body cross-section and the length of flight. The results of the field tests show that the air deflectors were not completely effective in reducing the amount of active ingredients dispersed in the air. The results of the static tests with drill equipped with the prototype indicated reductions of the active ingredient air concentrations ranging from 72 % up to 95 %, with reference to the conventional drill. Such ratios were applied to the amounts of active ingredients intercepted by the bees in the virtual tunnel contributing to a consistent reduction of the probability that sub-lethal effects can occur.