Metabolic competitive modes between digestion and locomotion are classified into three categories, termed the additive, digestion- and locomotion-priority modes. In nature, the caudal fin is frequently observed to sustain damage as a result of social rank, predation or disease. To test whether the metabolic mode changed differently for fish with different metabolic mode after caudal fin amputation as a consequence of intensified energy competition, we investigated the swimming performance of fasting and fed fish with and without the caudal fin in juveniles of three cyprinid fish species: qingbo (Spinibarbus sinensis, locomotion-priority mode), common carp (Cyprinus carpio, additive mode) and goldfish (Carassius auratus, digestion-priority mode). The critical swimming speed (Ucrit) of fasting qingbo, common carp and goldfish decreased significantly by 49%, 32% and 35% after caudal fin amputation. The maximum tail beat amplitude (TBAmax) (all three fishes), maximum tail beat frequency (TBFmax) (only common carp and goldfish) and (or) active metabolic rate (M˙O2active) (only common carp) increased significantly after caudal fin amputation. In the control fish, digestion let to a significantly lower Ucrit in goldfish but not in qingbo and common carp, and the M˙O2active of digesting common carp was higher than that of fasting fish, suggesting locomotion-priority, additive and digestion-priority metabolic modes in qingbo, common carp and goldfish, respectively. However, after fin amputation, digestion showed no effect on Ucrit in any of the three fishes, and only the digesting common carp showed a higher M˙O2active than their fasting counterparts. This result suggested that the metabolic mode of the goldfish changed from the digestion- to the locomotion-priority mode, whereas the metabolic mode of the other two fishes remained the same after fin amputation. The metabolic mode of the common carp showed no change after fin amputation likely due to the high flexibility of the cardio-respiratory capacity of this fish, as indicated by the increased M˙O2active. Although the metabolic mode remained the same, the feeding metabolism in the fin-amputated qingbo was down-regulated at a lower swimming speed than that of the control group due to the intensified competition between digestion and locomotion. The underlying mechanism for the metabolic mode change in the goldfish is not clear and needs further investigation. However, we speculated that in caudal-fin-intact goldfish, the decreased swimming efficiency, rather than irreducible digestive loading, caused a decreased Ucrit in digesting fish (i.e. false digestion-priority mode), and the metabolic mode should not be judged simply by the relative magnitude of the metabolic rates of fasting and digesting fish.