We compared growth simulations by dynamic energy budget (DEB) and scope for growth (SFG) models of the Pacific oyster Crassostrea gigas, cultivated in Bourgneuf Bay on the French Atlantic coast. This bay is located at a latitude in the middle of the European range of the species, and is characterized by high concentrations of suspended particulate matter (SPM) and a marked gradient between high-turbidity sites in the north (daily SPM > 500 mg L-1) and intermediate-turbidity sites in the south. The models use two forcing variables: seawater temperature and food density. We tested two indices of food availability: chlorophyll a and microalgal concentrations. In the SFG model, food intake is simulated by a type-II Holling functional response, as in the DEB formulation, and the effect of turbidity in both models is therefore taken into account principally through the half-saturation coefficient for this functional response. Chlorophyll a concentrations were three to four times higher at the high-turbidity site, but oyster growth rates were significantly lower at this site than at the intermediate-turbidity site. A comparison of observed and simulated values showed that the DEB model performed better than the SFG model if microalgal concentration was used as an index of food availability, with the SFG model underestimating oyster growth in summer and autumn. However, the SFG model was much more efficient if chlorophyll a concentrations were used, with the DEB model systematically overestimating summer and autumn growth. This comparison suggests that both SFG and DEB simulations could be improved, to give a more accurate description of oyster growth in a turbid environment, and that the pre-ingestive selection mechanisms used by suspension feeders in turbid environments should probably be included in the formulation of feeding processes.