Three factors are examined as an aid to interpreting the historical sequence of Skeena sockeye catches. One of these (Exploitation Mechanism 1) affects the catch directly, the other two affect catch by way of increasing or decreasing the recruitment produced by a given number of spawners. "Exploitation Mechanism 1" of Ricker (1973) makes the catch taken from any given brood during a time of increasing rate of exploitation greater than when exploitation is steady or decreasing; this is the main reason for excellent catches from 1905 to 1925. "Exploitation Mechanism 2" makes the progeny of any given number of spawners less numerous under conditions of maximum sustainable yield than during the period when rate of utilization is increasing; as a result recruitments observed prior to 1930 overestimate the potential productivity of the system and hence the sustainable yield. Mechanisms 1 and 2 are both operative even if the best rate of exploitation is never exceeded. In fact, however, the rate of exploitation began to exceed the optimum level of 50–55% about 1908, continued to increase until 1928, and then slowly decreased until the optimum level was reached about 1945; during the latter period, and for about 15 yr subsequently, catches were further depressed as spawning stocks were slowly rebuilt. Another factor affecting recruitment is the apparent existence of interaction between neighboring year-classes in respect to survival rate. Although the exact mechanism of this interaction is not yet known, it has produced three sequences of moderately "dominant" year-classes, the most recent being 1959–63–67–71, in spite of the complication of having two important ages of maturity for Skeena sockeye. More important is the fact that during the recent period large spawnings (0.75–1 million) have produced poor returns while moderately small spawnings (0.25–0.5 million) have produced larger year-classes than they did at any earlier period — in spite of the handicap of Mechanism 2. This apparent interaction between broods implies that one or more small spawnings are a necessary prerequisite for unusually successful reproduction by a later large one, and that successful large spawnings depress the reproduction of subsequent smaller ones. When all these factors are considered together, the observed trend of the Skeena catches can be reproduced. Maximum natural sustainable yield from the system would have been achieved about 1960 except for 2 yr of much reduced reproduction caused by the Babine River rock slide of 1951. Currently, the increased numbers of fry produced in artificial spawning channels at Babine Lake have contributed to large smolt runs that are now beginning to appear in the adult catches.