The 'membrane pacemaker theory of metabolism' proposes that membranes set metabolic rate by modulating protein activity, and thus purports to explain membrane fatty acid allometry. This relationship has never been tested across species in ectotherms. After accounting for phylogeny, recent analyses have failed to support this theory based on correlations between muscle membrane composition and body mass across mammals. Therefore, the goal of this study was to seek phylogenetically corrected correlations between membrane composition, body mass and calcium-ATPase activity, using 12 species of closely related cypriniform fish (4-5500 g) covering a much narrower genetic scale than in previous tests. The results show that fish membrane unsaturation decreases with mass, but through different mechanisms from those in endotherms: 16:0 replacing 22:6 in muscle and 18:0 replacing 16:1, 18:1 and 18:2 in liver. This shows that allometric patterns differ between endotherms and ectotherms as well as between tissues. After accounting for phylogeny, however, almost all these relationships lose significance except for overall unsaturation. No relationship between calciumATPase activity and mass or phospholipid composition was detected. This study shows that membrane unsaturation of cypriniforms decreases with mass, but that genetic cues unrelated to size account for differences in the relative abundance of individual fatty acids. The membrane pacemaker concept accurately predicts general membrane properties such as unsaturation, but fails to explain finer scale allometric patterns. Future examinations of the membrane pacemaker hypothesis will have to take into account that allometric patterns vary between endotherms and ectotherms and between tissues of the same animal class.