Basic principles of fishery forecasts developing and their performance evaluation are considered for the modern forecasting of pacific salmon stocks dynamics in Kamchatka in 2018–2024. The forecasts validity is critically analyzed for the main units of the stocks, with attention to correct interpretation of the results and possible reasons of discrepancy between predicted and actual stock dynamics. Methodological aspects of forecasting are described taking into account the species specifics for pacific salmon. Certain stock units whose abundance forecasting is available using mathematical modeling are outlined. In the modern period of high abundance of salmon, a trend approach to forecasting is rational, because of possibility of prompt adjustment of the catch values in dependence on actual volume of the salmon runs. Climate and oceanographic conditions at Kamchatka are described for the modern period. Ecosystem nature of the discrepancy between predicted and actual rates of runs and catches for pacific salmon in 2020 and 2024 is shown. Density factor in the ecosystem due to significant increase of the salmon abundance in 2010–2020s is determined as the most likely reason of unexpectedly weak runs, presumably because of heightened mortality of fish with the length and weight decreasing under environmental changes negative for their food base. Special attention in forecasting should be paid to vast areas of positive or negative SST anomalies on the feeding grounds of Kamchatka herds of pacific salmon in the southwestern Bering Sea and North-West Pacific during their early pre-spawning migrations.