Previous laboratory studies reported disruption of the digestive physiology and learning behaviour in individual honeybees treated with Bowman-Birk inhibitor (BBI), a serine proteinase inhibitor expressed in some GM plants. Our objective was to detect behavioural effects of this transgene on honeybees at the colony level, maintained in laboratory conditions. We set up a choice experiment, based on 150 free-flying individuals which performed a over 7700 visits on the flowers. The mean number of visits per hour, the mean time spent on the feeder and the interval between consecutive visits were not significantly different when the feeding sucrose solution was mixed with BBI at 100 µg*mL-1, a dose close to the expression level in planta. The methodology proposed herein could form a colony scale procedure particularly relevant to the risk assessment of the impact on bees of proteinase inhibitors or other transgenes to be possibly expressed in melliferous plants.