Several measures have been developed to quantify swimming performance to understand various aspects of ecology and behaviour in fish species, as well as to help design functional applications for fishways and aquaculture. The aim of this study was (1) to explore the relevance between the optimal swimming speed (U_opt) and preferred swimming speed (U_pref), and (2) to investigate the relationship between locomotive energetic metabolism and flow velocity selection in aquatic organisms. For the study, we selected pale chub (Zacco platypus), a fish species typically occurring in riptide habitats, and Chinese bream (Parabramis pekinensis), a fish species inhabiting lentic ecosystems, as experimental models. In experiment I, both fish species were videotaped individually (N=13 for each species) in a self-made device to determine U_pref. The device consisted of a 1-m-long conical raceway with water speed ranging from 102.7 to 18.6 cm/s and water temperature of (25±0.5)℃. The raceway was artificially divided into 5 flow velocity intervals with equal length (i.e. 20 cm each intervals). The videos were then analyzed by Ethovision XT19. The percentage of time stayed in each flow velocity interval during the whole video period (P_t) and frequency of crossing each intervals (P_f) were calculated. In experiment II, the critical swimming speed (U_crit) and swimming metabolic rate (M_O2) were measured at different swimming speeds. Resting metabolic rate (RMR), cost of transport (COT), net cost of transport (COT_net), and U_opt were calculated for each species. The experimental results were as follows: (1) both U_crit and RMR of Chinese bream were significantly higher than those of pale chub (P < 0.05), whereas maximum metabolic rate (MMR) showed no significant difference between two fish species. (2) M_O2 increased significantly while COT decreased significantly with swimming speed in both fish species. Furthermore, Chinese bream showed lower M_O2 and COT values within low swimming speed range while it showed higher M_O2 and COT values within high swimming speed range compared to those of pale chub. (3) COT_net of the two fish species increased with swimming speed, and the difference of COT_net (pale chub showed lower COT_net than Chinese bream) between the two species gradually increased with the increase in swimming speed (P < 0.05). (4) U_opt of Chinese bream ((6.20±1.29) body length (BL)/s) was significantly higher than that of pale chub ((11.56±1.57) BL/s). (5) P_t in the area of the minimum velocity interval (18.6-23.8 cm/s) in the present study was significantly larger than those in other areas (P < 0.05), suggesting that U_perf of both fish species was lower than (or equal to) the range 18.6-23.8 cm/s. However, P_t of Chinese bream in the area of the minimum velocity interval was significantly larger than that of the pale chub (P < 0.05), whereas P_t of Chinese bream in the areas of high velocity intervals was significantly lower than that of the pale chub. The present study demonstrated that U_perf and U_opt were irrelevant in both fish species. Pale chub, which typically occur in torrent habitats, preferred higher flow velocity; this may be related to its locomotive energetic metabolism such as high U_opt and U_crit and low COT_net, when compared to Chinese bream.