Understanding the causes of water quality change is important for managing water resources. Yet it is often difficult to identify and quantify the primary drivers of changes in riverine concentrations and loads. On a national scale in particular, cohesive assessments of water quality can provide important insights into changes in water availability, but driver attribution is challenging due to the wide range of watershed and climate conditions which must be considered. When considering nutrients specifically, concentrations and loads have multiple sources and sinks, and interactions within a watershed affect their fate and transport. This presentation focuses on a new approach that exploits seasonally dynamic nutrient source estimates to evaluate the primary drivers of seasonal and annual trends in observed total nitrogen (TN) and total phosphorus (TP) concentrations and loads in rivers that drain to coastal waters across the U.S. Recent advances in spatially referenced regression on watershed attributes (SPARROW) modeling include seasonally dynamic predictions of nutrient loads and sources between 2000-2020 for watersheds across the U.S. Trends in observed TN and TP loads were estimated in over 20 coastal rivers across the U.S. were determined for the 2000-2020 period using the Weighted Regression on Time, Discharge, and Season (WRTDS) model. Predictions from these two models allow for direct comparisons between observed nutrient trends and changes in upstream watersheds, as well as estimates of trends in unmonitored streams. Modeled timeseries of seasonal and annual nutrient source loads from contributing watersheds will be compared to seasonal and annual trends in nutrients at monitoring sites to evaluate how changes in nutrient sources may contribute to changing TN and TP loads to coastal waters.