The Pacific oyster, Crassostrea gigas, is one of the most important global aquaculture species due to its potential high growth rates and tolerance of a wide range of environmental conditions; however, most farmers use seed from wild, non-domesticated stocks for production. In this study we estimated genetic parameters for performance traits of oysters from the Molluscan Broodstock Program (MBP), a family-based selective breeding program designed to improve yields of Pacific oysters on the West Coast, USA. Covariance components were obtained using AIREMLF90, and a multiple-trait animal model. A total of 15,236 records were analyzed from five generations of selectively bred oysters. Heritability estimates for field traits at harvest were all positive over the five analyzed generations, ranging from 0.58±0.03 (3rd generation) to 0.30±0.04 (5th generation) for yield, from 0.55±0.03 (1st generation) to 0.12±0.02 (4th generation) for survival, and from 0.51±0.03 (2nd generation) to 0.40±0.03 (4th generation) for mean individual weight at harvest (growth). Declines in heritabilities for survival and yield in the 4th and 5th generations were perhaps a result of changing ocean conditions due to increased upwelled hypoxic and acidified seawater occurring on the West Coast, USA, that affected the quality of seed from the MBP hatchery/nursery. Realized heritabilities were all positive and medium-to-high across generations, ranging to 0.11±0.42 (for survival in the 3rd generation) to 1.20±0.35 (for individual weight in the 5th generation); however, standard errors for realized heritabilities were high (ranging 0.15 to 1.47), especially in the 5th generation. There was a gradual improvement in genetic gains for survival and yield over the five generations of selection, resulting in accumulated gains of +15.7% and +19%, respectively, in the 5th generation; however, little improvement was achieved after the 2nd generation in genetic gain for individual weight, with an accumulated gain in the 5th generation of +11.3%. Realized gains for performance traits were less than predicted by genetic gains. Realized gain in survival consistently improved over the selection period, resulting in an accumulated gain of 11.7% in the 5th generation compared to that of wild, non-selected controls; however, gains in individual weight and yield at harvest were not consistent across generations and gains in the 5th generation were −9.8% and 0%, respectively, compared with those of controls, perhaps due to inconsistent genetic quality of control broodstock sampled from wild populations. Across generations, there were positive genetic correlations between yield and both survival (0.38±0.04) and individual weight (0.90±0.01) as well as between survival and individual weight (0.25±0.04). Positive medium-to-high genetic correlations among harvest traits suggest that indirect gains in yield can be achieved by selection for either higher growth or survival. Statement of relevanceThe Pacific oyster is an important aquaculture species both globally and on the West Coast, US. This study describes long-term (over 5 generations) genetic and phenotypic responses of Pacific oysters planted at commercial farms, in response to selection for yield, survival and mean individual weight at harvest. The breeding program spanned a period of increased ocean upwelling and acidification on the West Coast, US, that may have reduced heritabilities for survival due to impacts of the quality of hatchery seed. Results will be of interest to other breeding programs for Pacific oysters in France, Australia and New Zealand, as well as breeders of other extensively farmed aquaculture species. Genetically improved broodstock will help support future oyster production under more challenging conditions, due to the global effects of acidification and warming of coastal waters.