Comparative physiologists and ecologists have searched for a specific morphological, physiological or biochemical parameter that could be easily measured in a captive, frozen, or preserved animal, and that would accurately predict the routine behavior or performance of that species in the wild. Many investigators have measured the activity of specific enzymes in the locomotor musculature of marine fishes, generally assuming that high specific activities of enzymes involved in aerobic metabolism are indicators of high levels of sustained swimming performance and that high activities of anaerobic metabolic enzymes indicate high levels of burst swimming performance. We review the data that support this hypothesis and describe two recent studies we have conducted that specifically test the hypothesis that biochemical indices of anaerobic or aerobic capacity in fish myotomal muscle correlate with direct measures of swimming performance. First, we determined that the maximum speed during escapes (C-starts) for individual larval and juvenile California halibut did not correlate with the activity of the enzyme lactate dehydrogenase, an index of anaerobic capacity, in the myotomal muscle, when the effects of fish size are factored out using residuals analysis. Second, we found that none of three aerobic capacity indices (citrate synthase activity, 3-hydroxy-o-acylCoA dehydrogenase activity, and myoglobin concentration) measured in the slow, oxidative muscle of juvenile scombrid fishes correlated significantly with maximum sustained speed. Thus, there was little correspondence between specific biochemical characteristics of the locomotor muscle of individual fish and whole animal swimming performance. However, it may be possible to identify biochemical indices that are accurate predictors of animal performance in phylogenetically based studies designed to separate out the effects of body size, temperature, and ontogenetic stage.