The sustained swimming speeds of juvenile coho and chinook salmon tested in a tubular chamber at temperatures ranging from 10 to 20° C. usually were dependent on the dissolved oxygen concentration at any tried concentration below the air‐saturation level. Some underyearling chinook salmon tested at 19.5° C. were notably exceptional; their performance at high oxygen concentrations was not markedly better than that observed at lower concentrations between 5 and 7 mg/1. These fish may have been anemic. Reduction of oxygen concentration from air‐saturation levels to 7, 6, 5, 4, and 3 mg/1 usually resulted in reduction of the maximum sustained swimming speed of coho salmon by about 5, 8, 13, 20, and 30 percent, respectively. These percentages are based on the observed performance of the first‐failing and second‐failing individuals of a group of five fish used in a performance test. The corresponding estimated percent reductions of the swimming speed of first‐failing fish in most of the early exploratory experiments with chinook salmon were somewhat greater, averaging approximately 10, 14, 20, 27, and 38 percent, respectively. Within the ranges of tested concentrations above the saturation levels, variations of oxygen concentration had little or no effect on the performance of wild underyearling and hatchery‐reared yearling coho salmon. The performance of the wild underyearlings at tested oxygen concentrations improved progressively with increase of temperature from 10 to 15 and to 20° C.