Healthy river ecosystems are conducive to the sustainable use of water resources. As the climate change and human activities become more intense, the integrity of such ecosystems is seriously threatened. The maintenance of enough environmental flow (e‐flow) for river ecosystems is the most effective way of protecting river health. Accurately and reasonably estimating e‐flow is essential for river health maintenance. So far, few methods can quantitatively verify the e‐flow results calculated by various e‐flow methods, which reduces the success rate of ecological restoration projects on a global scale. Therefore, the Tennant method, wetted perimeter (WP) method, and adapted ecological hydraulic radius (AEHRA) method have been recognized by many scholars. The article used the combination of these three methods to calculate the e‐flow and its meeting rate (actual flow/e‐flow) and formed a new framework to verify e‐flow results. First, the Shannon diversity index, index of biological integrity (IBI), and river health index are used to evaluate river health status. Then, we studied the relationship between e‐flow meeting rates and three indices using field monitoring data on hydrology, water quality, and biological communities. Finally, we investigated the effects of different e‐flow calculation methods on river health. Results show that the e‐flow in the centre of the study area calculated by the Tennant and WP methods are relatively high (48.33–317 m3/s), whereas lower e‐flow (0.03–21 m3/s) are observed in its southern and northern parts. The e‐flow in the southern mountain area calculated by the AEHRA method (1.42 m3/s) is higher than those in the other regions determined by the same method. Furthermore, the highest values of river health—the Shannon diversity index (3.19), IBI (67.47), and river health index (0.65)—appear in mountainous areas that are less affected by human activities. The lowest values appear in the urban areas with river health values (0.97, 14.54, and 0.45, respectively) with high population density. For the river health scores, the values of the Shannon diversity index and IBI first increase and then decrease with the increase in the e‐flow meeting rates. The values of river health index continuously increase with the increase in the WP‐ and AEHRA‐calculated e‐flow meeting rate and fluctuate with the increase of Tennant‐calculated e‐flow meeting rate. We concluded that the adoption of WP‐ and AEHRA‐calculated e‐flow values can maintain the health of river ecosystems with a certain degree of pollution, whereas the Tennant‐calculated e‐flow can only ensure the health of river ecosystems with little pollution, the reason for which is the pollution‐affected relationships between river biota communities and stream flows, and the Tennant method uses historical stream flow records to calculate e‐flow. WP and AEHRA methods, on the other hand, do not rely on flow records. The methodologies and results can provide a scientific basis for the selection of suitable e‐flow methods and therefore facilitate the projects for ecological river restoration.