Differential survival and growth were tested in Chinook salmon Oncorhynchus tshawytscha expressing two common alleles, *–100 and *–260, at the superoxide dismutase locus (sSOD-1*). These tests were necessary to support separate studies in which the two alleles were used as genetic marks under the assumption of mark neutrality. Heterozygous adults were used to produce progeny with –100/–100, –100/–260, and –260/–260 genotypes that were reared in two natural streams and two hatcheries in the states of Washington and Oregon. The latter also were evaluated as returning adults. In general, the genotype ratios of juveniles reared at hatcheries were consistent with high survival and little or no differential survival in the hatchery. Adult returns at one hatchery were significantly different from the expected proportions, and the survival of the –260/–260 genotype was 0.56–0.89 times that of the –100/–100 genotype over four year-classes. Adult returns at a second hatchery (one year-class) were similar but not statistically significant: survival of the –260/–260genotype relative to the –100/–100 genotype was 0.76. The performance of the heterozygote group was intermediate at both hatcheries. Significant differences in growth were rarely observed among hatchery fish (one year-class of juveniles and one age-class of adult males) but were consistent with greater performance for the –100/–100 genotype. Results from two groups of juveniles reared in streams (one year-class from each stream) suggested few differences in growth, but the observed genotype ratios were significantly different from the expected ratios in one stream. Those differences were consistent with the adult data; survival for the –260/–260 genotype was 76% of that of the –100/–100 genotype. These results, which indicate nonneutrality among sSOD-1* genotypes, caused us to modify our related studies and suggest caution in the interpretation of results and analyses in which allozyme marks are assumed to be neutral.