We present statistical attribution of peak-flow trends for Northeast streams and rivers during the last 75 years, based on peak-flow data at 125 U.S. Geological Survey gages with significant trends. We used event precipitation just prior to all annual peak flows (from Global Historical Climatology Network stations near each gage), soil moisture in the month prior to each peak (Palmer Drought Severity Index), and basin developed area and major-reservoir storage for attribution testing. Our attributions were based on the interannual correlation of peak flows with event precipitation and soil moisture, and the consistency of peak-flow trends with trends in the potential causal variables. Comparison of peak-flow changes to event-precipitation changes often indicated influence of human basin alterations. For example, if precipitation increased and peak-flows decreased in a basin with large amounts of reservoir storage, it’s likely that reservoir storage additions caused the peak-flow decreases. In contrast, peak-flow changes that were larger than precipitation changes, in a basin with high amounts of development, indicate that basin urbanization likely affected the peak-flow changes.