This paper presents a power-law model of Shreve's link magnitude for estimating flow regime and the mean annual discharge in river networks. The model takes into account the effects of climatic, geological and land use conditions within river basins. High predictability of the model for discharge estimation is confirmed by using discharge observation data from 109 large river basins in Japan. The results indicate that the effect of precipitation on flow regime estimation is apparent for high flow discharges, while that of geology become obvious for low discharges or a drought condition. As for the land use types, the results show that wastelands and city areas have large impacts on the mean annual discharge. Since the link magnitude system with climatic, geological and land use conditions provides flow regime information throughout river networks, it could be very useful for basin-wide environmental management.