Burst swimming performance is an important component to characterize the survival fitness of fish. Habitat fragmentation poses a serious threat to the survival and reproduction of stream fishes, especially for migratory fishes, and places higher demands on the burst swimming performance of fish. Brachymystax tsinlingensis, a rare and endemic species in Qinling Mountain region, is a second-class state-protected wild animal in China Red Data Book of Endangered Animals. It is an endangered teleost fish species that belongs to the family Salmonidae and is generally regarded as one of the two southernmost-distributed Salmonid fish worldwide. B. tsinlingensis strongly prefers swift currents and clear water over a large gravel substrate. It presents migratory behavior that can be triggered by changes in seasonal temperatures. Importantly, B. tsinlingensis often needs to perform repeated burst swimming movements to cross the rapids or even dam barriers, so the burst swimming ability and its consistency are crucial for the survival, reproduction and routine activities of this species. Compared with the remarkable migratory habit of B. tsinlingensis, Phoxinus lagowskii, the most common sympatric species in their distribution area, tends to be sedentary. Hence, these two species are not only a superior animal model for studying the adaptive evolution theory of sympatric species, but also highly desirable subject targets in fish habitat conservation and restoration practices. In order to investigate the interspecific differences and intra-species variation in burst swimming performance of B. tsinlingensis and its sympatric species (i.e. P. lagowskii), the absolute burst swimming speed (U_burst) and relative burst swimming speed (rU_burst) as well as their repeatability of the two experimental fish at different life history stages were measured using a self-made ecologically simulated apparatus for the determination of fish burst swimming ability. The study was expected to provide a theoretical basis for biodiversity conservation in the Qinling stream ecosystem and future ecological restoration of B. tsinlingensis habitats. The results showed that: (1) the burst swimming performance (in terms of U_burst and rU_burst) of both two experimental fish species had high repeatability (ICC coefficient ＞ 0.75), but the burst swimming performance of B. tsinlingensis had stronger resilience. (2) Overall, the U_burst of B. tsinlingensis was higher than that of P. lagowskii, and the difference in U_burst between the two was significant; thus, there was no convergent adaptation between the two sympatric species. (3) Life history stage had significant effects on U_burst and rU_burst in both B. tsinlingensis and P. lagowskii, and the patterns of intraspecific variation in burst swimming ability were similar in both species, with U_burst increasing while rU_burst decreasing with the increasing development stages. The findings suggest that future fish passage design or wading project evaluation related to fish swimming ability should integrate life history stage effects at the intraspecific level as well as differences in species ecological habits and performances at the interspecific level.