Dependence of chum salmon Oncorhynchus keta recruitment (number of mature adults) in Okhotsky district located on the Okhotsk Sea coast on a number of spawners is analyzed for the spawning years of 1983–2014. During this period, the progeny abundance increased, then stabilized on a high level, then decreased. Generally, the number of recruits depended poorly on the number of spawners because of clear environmental impact. Indeed, the recruitment variation could be simulated on the base of Ricker curve with addition of a variable responsible for environmental trend. Number of the year of spawning was used as such variable. So, the chum recruitment R is modeled as R = a. P. exp(–(P/b + ((X – c)/d)2 )) for the period of increasing and R = a. P. exp(–(P/b – ((X – c)/с)2 )) for the period of decreasing, where P — number of parents, X — year of spawning, and a, b, c, d — empirical coefficients. This model explains > 80 % of the recruitment deviation. The environmental impact was weak in the period of stable recruitment, so the direct dependence of progeny abundance on spawners number (Ricker curve) can be used for these years, with similar determination (r2 > 0.8). This approach is proposed as a possible compromise between hypotheses on climate-driven and autoregulating mechanisms of the salmons abundance variability.