AbstractMany river systems in western North America retain a fluvial strath-terrace rec ord of discontinuous downcutting into bedrock through the Quaternary. Their importance lies in their use to interpret climatic events in the headwaters and to determine long-term incision rates. Terrace formation has been ascribed to changes in sediment supply and/or water discharge produced by late Quaternary climatic fluctuations. We use a one-dimensional channel- evolution model to explore whether temporal variations in sediment and water discharge can generate terrace sequences. The model includes sediment transport, vertical bedrock erosion limited by alluvial cover, and lateral valley-wall erosion. We set limits on our modeling by using data collected from the terraced Wind River basin. Two types of experiments were performed: constant- period sinusoidal input histories and variable-period inputs scaled by the marine δ18O rec ord. Our simulations indicate that strath-terrace formation requires input variability that produces a changing ratio of vertical to lateral erosion rates. Straths are cut when the channel floor is protected from erosion by sediment and are abandoned—and terraces formed—when incision can resume following sediment-cover thinning. High sediment supply promotes wide valley floors that are abandoned as sediment supply decreases. In contrast, wide valleys are promoted by low effective water discharge and are abandoned as discharge increases. Widening of the valley floors that become terraces occurs over many thousands of years. The transition from valley widening to downcutting and terrace creation occurs in response to subtle input changes affecting local divergence of sediment-transport capacity. Formation of terraces lags by several thousand years the input changes that cause their formation.Our results suggest that use of terrace ages to set limits on the timing of a specific event must be done with the knowledge that the system can take thousands of years to respond to a perturbation. The incision rate calculated in the field from the lowest terrace in these systems will likely be higher than the rate calculated by using older terraces, because the most recent fluvial response in the field is commonly downcutting associated with declining sediment input since the Last Glacial Maximum. This apparent increase in incision rates is observed in many river systems and should not necessarily be interpreted as a response to an increase in rock-uplift rate.