The timing of breeding migration and reproduction links generations and substantially influences individual fitness. In salmonid fishes, such phenological events (seasonal return to freshwater and spawning) vary among populations but are consistent among years, indicating local adaptation in these traits to prevailing environmental conditions. Changing reproductive phenology has been observed in many populations of Atlantic and Pacific salmon and is sometimes attributed to adaptive responses to climate change. The sockeye salmon spawning in the Cedar River near Seattle, Washington, USA, have displayed dramatic changes in spawning timing over the past 50 years, trending later through the early 1990s, and becoming earlier since then. We explored the patterns and drivers of these changes using generalized linear models and mathematical simulations to identify possible environmental correlates of the changes, and test the alternative hypothesis that hatchery propagation caused inadvertent selection on timing. The trend toward later spawning prior to 1993 was partially explained by environmental changes, but the rapid advance in spawning since was not. Instead, since its initiation in 1991, the hatchery has, on average, selected for earlier spawning, and, depending on trait heritability, could have advanced spawning by 1–3 weeks over this period. We estimated heritability of spawning date to be high (h 2 ~0.8; 95% CI: 0.5–1.1), so the upper end of this range is not improbable, though at lower heritabilities a smaller effect would be expected. The lower reproductive success of early spawners and relatively low survival of early emerging juveniles observed in recent years suggest that artificial and natural selection are acting in opposite directions. The fitness costs of early spawning may be exacerbated by future warming; thus, the artificially advanced phenology could reduce the population’s productivity. Such artificial selection is known in many salmon hatcheries, so there are broad consequences for the productivity of wild populations comingled with hatchery‐produced fish.