Sea stars are one of the primary predators of shellfish and often cause mass mortality among cultured shellfish. To develop effective control strategies, it is critical to understand the feeding ecophysiology (e.g., feeding rate and prey selection) of sea stars. Asterias amurensis and Asterina pectinifera are the dominant sea star species in the coastal waters of China. We evaluated prey selection and the feeding rate of these two sea stars on three species of bivalves: scallops Chlamys farreri, clams Ruditapes philippinarum, and blue mussels Mytilus galloprovincialis. The experimental sea stars and bivalves were collected from Sanggou Bay, Northern China and transported to our seaside laboratory. The animals were acclimated to laboratory conditions for 7 d prior to the initiation of the experiment. The experiment was conducted between March 19 and April 5, 2008, at different seawater temperatures. Following acclimation, the sea stars were placed in cement tanks (L×W×H=6×1×0.5 m) at a density of 5 individuals per tank (single species per tank). The sea water was pumped from Sanggou Bay and sand filtered. Daily water exchange was ca 100%. We placed the bivalves (N=30 per species) evenly in each tank to ensure the sea stars had an equal probability of encountering the three species of bivalves. Each treatment was conducted in triplicate (N= 3 tanks). The number of bivalves of each species preyed upon by the sea stars was recorded twice daily at 7:00 and 17:00. In addition, we measured the Q₁₀ coefficient at water temperatures ranging from 4.3-7.8℃ and from 7.8-13.3℃. Both species of sea star preyed on all three bivalve species. Similarly, both species exhibited preference in the order clam>mussel>scallop. During the first part of the experiment (March 19-24), A. amurensis preyed on 64.28, 28.57, and 7.14% of the scallops, clams, and blue mussels, respectively. A. pectinifera preyed upon 50, 44.44, and 5.56% of the bivalve species, respectively. The mean feeding rates of A. amurensis and A. pectinifera on the clam (0.50 and 0.37 ind/d, respectively) and blue mussel (0.26 and 0.27 ind/d, respectively) were significantly higher than those on the scallop (0.05 and 0.07 ind/d, respectively). The feeding rate was significantly influenced by water temperature and generally increased with increasing water temperature. The total mean feeding rates of the two sea stars were 0.69 and 0.79 g·d^-1, respectively (based on dry tissue weight of bivalves). As water temperature increased from 4.3 to 7.8℃, the Q₁₀coefficients for A. amurensis and A. pectinifera were 6.38 and 2.33, respectively. However, when the water temperature was increased from 7.8 to 13.3℃, there was no increase in the feeding rate (Q₁₀=1.13 and 1.22, respectively). Our results have implications for the provision of protective refuges for the species of interest (i.e., scallops) during culture in suspended lantern nets. Protective strategies are most likely to be needed when the water temperature increases above 4.3℃, as the feeding rate and activity of sea stars increased significantly above this point. Based on prey selectivity, bivalves that have a lower commercial value (e.g., blue mussels) may be co-cultured in the scallop lantern nets to serve as a buffer against predators. Furthermore, any sea stars present in the cultivation nets should be removed during the buffering period.