The escapement of sockeye salmon, Oncorhynchus nerka, into the Kenai River has been estimated using a combination of sonar and fish wheels operated at rkm 31.1 since 1968. Beginning in 2011, the fisheries were managed using a dual-frequency identification sonar (DIDSON). This project used mark–recapture methods to independently investigate bias in DIDSON estimates of sockeye salmon escapement in 2006–2008. Mark–recapture experiments were conducted using passive integrated transponder tags, radio tags, fish wheels, and weirs (Russian River and Hidden Creek) as recapture sites. Radiotelemetry was used primarily to identify spawning areas throughout the watershed and estimate tagging-induced mortality. Overall, our mark–recapture abundance estimates did not support the conclusion that the DIDSON estimates were consistently biased during the 3 years of this study. In 2006, the Russian mark–recapture estimate was significantly higher while the Hidden mark–recapture estimate was significantly lower than the DIDSON estimate. Variable tagging-induced mortality or initial capture probabilities between the 2 stocks were likely causes of these observed differences. In 2007, the mark–recapture estimates from both weirs were not different from the DIDSON estimate. However in 2008, the Russian mark–recapture estimate was not different while the Hidden mark–recapture estimate was significantly lower than the DIDSON estimate. A linear regression of the pooled mark–recapture estimates (dependent variable) against the DIDSON estimates (2006–2008) resulted in an intercept not significantly different from zero and a slope not significantly different from one. Radiotelemetry studies indicated that 35–42% of sockeye salmon spawned in the mainstem Kenai River between the Russian River confluence and Skilak Lake. Another 10–20% spawned in an approximately 16 km segment of the Kenai River immediately below Skilak Lake, while 11–21% spawned in upper tributaries of the watershed (Russian River, Quartz Creek, Trail Lakes, Ptarmigan Creek, and others).