The swept-area estimates of biomass from the triennial groundfish trawl surveys on the shelf of the U.S. West Coast are believed to seriously underestimate rockfish (Scorpaenidae) stock biomasses. The bulk catchability (Q), defined to be the ratio between swept-area biomass and actual biomass, is herein modeled using a Bayesian age-structured meta-analysis of suitable West Coast rockfish stocks. Six shelf stocks of rockfish were used. The posterior distribution of Q was insensitive to choice of prior and gives a probability of about 0.05 that the bulk catchability of a randomly selected shelf rockfish species will be unity or higher. Between survey variability in bulk catchabilities was modeled as a multiplicative main effect. With individual posterior probabilities in excess of 0.99, bulk catchabilities were lower than normal in 1977 and 1980 and higher than normal in 1989 and 1998. The low catchabilities in 1977 and 1980 are consistent with previously identified problems with lack of bottom contact in the earlier years of the survey. Future work will extend the model to incorporate dynamic modeling of unassessed rockfish stocks, and suggestions for this are given.