The article presents a strategy for classifying streamflows into groups based on their temporal dynamics. It compares dynamic patterns of river discharges over time, capturing seasonality and short-term components using linear time series models. The final classification offers a means of identifying hydrological regimes and gives a preliminary understanding of the potential impacts of climate change on streamflows. The approach is illustrated by analyzing streamflow data from 221 stations in the United States in two nonoverlapping periods.