Estimating fishery harvest and spawning escapement (spawning stock size) are critical components of fisheries management; however, they can be particularly challenging to measure in systems where visually indistinguishable, but reproductively isolated populations mix within a single fishing area. Genetic stock identification is a common tool used in such mixed‐stock fisheries to improve estimates of spawning escapement and productivity; however, there are few references for management applications, particularly for inland recreational fisheries. The kokanee Oncorhynchus nerka population in Wood Lake, British Columbia, is a highly productive and valuable mixed‐stock fishery that includes two reproductively distinct ecotypes: shore‐ and stream‐spawning. Enumeration of shore‐spawning kokanee is logistically challenging, as the spawning population is not confined to a defined area or depth like stream‐spawners. Here, we combined in‐lake sampling (angler harvest and age‐0 trawl samples) over a 9‐year period (2008–2016) with genetic stock identification and Bayesian statistics to develop a new method for enumerating shore‐spawning kokanee. Our results suggest that angler‐harvested kokanee are representative of the spawner age structure and stock proportions. Therefore, we used the angler harvest sample combined with known stream‐spawner escapement to reconstruct the shore‐spawner escapement time series. Shore‐spawner abundance varied between 2,040 spawners and 13,460 spawners across years, which is over four times that previously predicted using the peak estimate of visual survey counts. Our results demonstrate the recovery of both the shore‐ and stream‐spawning kokanee in Wood Lake following a well‐documented crash in 2011 and suggest that a larger harvestable surplus is available for this high‐value kokanee fishery.