The impacts of pesticides, and in particular of neonicotinoids, on bee health remain much debated. Many studies describing negative effects have been criticized as the experimental protocol did not perfectly simulate real-life field scenarios. Here, we placed free-flying bumblebee colonies next to raspberry crops that were either untreated or treated with the neonicotinoid thiacloprid as part of normal farming practice. Colonies were exposed to the raspberry crops for a two week period before being relocated to either a flower-rich or flower-poor site. Overall, exposed colonies were more likely to die prematurely, and those that survived reached a lower final weight and produced 46% fewer reproductives than colonies placed at control farms. The impact was more marked at the flower-rich site (all colonies performed poorly at the flower poor site). Analysis of nectar and pollen stores from bumblebee colonies placed at the same raspberry farms revealed thiacloprid residues of up to 771 ppb in pollen and up to 561 ppb in nectar. The image of thiacloprid as a relatively benign neonicotinoid should now be questioned.