This study examines the possibility that wind mixing in Shelikof Strait, Gulf of Alaska, is a critical factor for larvae of walleye pollock Theragra chalcogramma. The abundances of walleye pollock larvae hatched on a given day and surviving through the early feeding stage were determined by in situ sampling and otolith analysis for 1983 and 1985 to 1991. Periods of anomalously low or high larval survival were determined by comparing observed first-feeding date distributions of survivors sampled in late May surveys with expected first-feeding date distributions from a model utilizing information on spawning time and abundance, measured egg mortality, assumed larval mortality, and survey date. The cube of the wind speed represented daily estimates of mixing for the same years; wind speeds were determined from gridded sea-level pressure data using a geotriptic wind model. When the resulting daily distributions of larval abundance and mixing were compared, 2 patterns emerged: (1) strong wind mixing events during the first-feeding period were associated with periods of lower than expected larval survival, and (2) periods of higher than expected larval survival were associated with calm with periods often bracketed by strong mixing. The results indicate that over the 8 yr of observation strong mixing events during the first-feeding period were detrimental to survival of pollock larvae.