A computer-implemented method for dynamically adjusting rendering parameters based on user movements may include determining viewpoint movement data for a user viewing a rendering of a 3D model at a first time, determining a first level-of-detail at which to render the 3D model based at least in part on the viewpoint movement data at the first time and rendering the 3D model at the first level-of-detail. The method may also include determining viewpoint movement data for the user at a second time, wherein the viewpoint movement data at the second time differs from the viewpoint movement data at the first time. In addition, the method may include determining a second level-of-detail at which to render the 3D model based at least in part on the viewpoint movement data at the second time and rendering the 3D model at the second level-of-detail, wherein the second level-of-detail differs from the first level-of-detail.