- Samples, consisting of scales, length and weight measurements, and sex determinations of chum, pink, and silver salmons, were taken from the commercial catch in the Columbia River in 1914. Five hundred eighteen chum scales were examined. All fish had gone to sea early in their first year; and 70.5 percent had returned in their third year, 28.7 percent in their fourth, and 0.8 percent in their fifth. Growth curves, based on scale and fish measurements, are given for all sex-age groups. The length-weight relationship may be expressed by the equation Weight = 0.0001378 x Length [superscript 2.6004]. Tendencies were noted, during the course of the run, for older fish and for males to be relatively more abundant in the earlier portions of the run. In comparisons with data from other localities, it was demonstrated that, from south to north: (1) there is a decrease in mean length at the same age; (2) older fish are progressively more abundant; and (3) the runs are progressively earlier. Pink salmon are uncommon in the Columbia River, and only six were examined. All had gone to sea early in their first year, and all had returned in their second year. Eight hundred eighty-five silver salmon scales were examined. Of the total, 6.1 percent had gone to sea early in their second year, and returned late in their second year (2 [subscript 2]); 0.3 percent had gone to sea early in their third, and returned late in their third (3 [subscript 3]); 83.9 percent had gone to sea in their second, and returned in their third (3 [subscript 2]); and 9.7 percent had gone to sea in their third year, and returned during their fourth year (4 [subscript 3]). Growth curves are given for all, age groups. The length-weight relationship may be expressed by the equation Weight = 0.0000373 x Length [superscript 2.8945]. As in the case of the chums, changes were noted during the run in the sex ratio and the percentage of age classes. Comparison with data from other localities revealed south-to-north changes in size and in abundance of different age groups, similar to those demonstrated for the chums. It is felt that the differences between races of a single river system will prove to be of the same nature as those here demonstrated between populations of different river systems. Knowledge of the differences between the races of salmons entering the Columbia River may be of extreme importance in the management of the fisheries.