Abstract.-Stock status of the blue crab Callinectes sapidus in Chesapeake Bay is determined by comparing current estimates of fishing mortality with biological reference points. Given the recent focus on blue crab conservation, there is a need to obtain reliable estimates of survival to compare with the biological reference points. A tagging program was initiated on the terminally molted, mature female component of the Chesapeake Bay blue crab stock to estimate annual and semiannual survival rates. Crabs were obtained from fishery-independent research surveys throughout the bay and were measured, tagged, and released on-site. Tagging was conducted twice per year (winter and summer) from November 2001 to March 2005. Annual survival and tag recovery rates were estimated independently for the winter and summer tagging data using Brownie models. The two independent estimates of annual survival based on winter tagging (mean 6 SE ¼ 0.081 6 0.031) and summer tagging (0.080 6 0.024) data were virtually identical and low. The estimated tag recovery rate was 24% based on winter tagging data and 17% based on summer tagging data. The estimated monthly survival rate during winter (mean 6 SE ¼ 0.868 6 0.016) was higher than that during summer (0.748 6 0.019). The low estimates of annual survival are consistent with (1) historical estimates of the percentage of age-2 and older females in the winter dredge fishery, and (2) recent estimates of exploitation rate obtained from the ratio of catch to initial abundance. These findings indicate that survival rates of mature female blue crabs in Chesapeake Bay have remained extremely low during a period of depressed abundance, which may be preventing stock recovery. Moreover, this study represents one of the few to derive experimental Brownie model estimates of semiannual survival of an invertebrate species subject to a continuous fishery.