The Crystal River site (8CI1) in west-central Florida is famous as the southernmost major participant in the Hopewell Interaction Sphere, and certainly has the most Hopewellian goods of any Woodland site in Florida. Sharon Goad (1978), among others, proposed that Crystal River secured this position by controlling the production and exchange of marine shell ornaments and cups. I test this hypothesis through the analysis of marine shell recovered from previous excavations, recent surface finds, and shell debris from 58 core samples extracted from the Crystal River mounds, plaza, middens, and surrounding marshland. The analysis reveals an abundance of shell ornaments in burials, but only a limited presence of marine shell used in ornament production around the site, which contradicts Goad's original hypothesis. Therefore, I propose several alternative explanations for the disproportionate presence of Hopewellian items at Crystal River.