There is little in the literature about the risks of manual handling of material in supermarkets and what there is refers solely to storehouse work. This contrasts with the substantial number of studies of the risk of repeated arm movements among supermarket cash-desk staff. The scarcity of information is partly due to the difficulties of applying widely employed, standardized evaluation methods in this sector. One of the conditions limiting the application of the NIOSH protocol in this retail sector is that lifting tasks are so often closely tied to transport. The biomechanical analysis method we used brought to light considerable risks in many of the steps investigated: unpacking the pallet, unloading the crates from the pallet to the ground, lifting them from the floor onto display stands, and filling the boxes on the stands with goods before the shop opens. Images acquired on site were analyzed in the laboratory. We selected the most indicative images, which were then studied as regards posture and biomechanics using Apalys 3.0 software (ILMCAD GmbH, Ilmenau, Germany). Biomechemical analysis was done on the following movements: unloading crates from the pallet, positioning them on fruit and vegetable department display stands, and filling the boxes on the stands. We obtained a prediction of 2720 to 5472 N for the load at the lumbosacral junction (L5-S1). Simulation of the NIOSH index gave a value of 2.69 in the only case where the Waters protocol could be applied.