We assessed the importance of flow regime to the success of native and non‐native fish species by analysing winter/spring seining data collected from 1987 to 1997 on the resident fish communities of the lower Tuolumne River, California. The data were analysed using regression models to predict the percentage of non‐native fish at a site. The regression models included various combinations of the variables longitudinal location of the site, mean April/May stream discharge in the year of sampling, and mean April/May stream discharge in the previous year. Comparison of the models indicated that the best model included longitudinal location and stream discharge in the previous year. This model is consistent with the hypothesis that flow in the previous year differentially affects reproductive success of native and non‐native species and thus the resulting community sampled in the following winter/spring. A detrended correspondence analysis of percentage abundance species data identified a co‐occurring group of native species and a co‐occurring group of non‐native species with the non‐native red shiner (Cyprinella lutrensis) grouping separately. The differing reproductive strategies of the species were consistent with the hypothesis concerning spawning success. Our results indicate that flow regime is an important determinant of the reproductive success of native and non‐native fish species in regulated rivers. Manipulations of flow regime are a potentially powerful tool for managing native fish species, but should be considered in combination with other restoration efforts and in the context of ecosystem restoration.