 Precision farming, PF, management strategies are commonly based on estimates of within-field yield potential, which are usually derived from remotely sensed data such as vegetation index, 6, maps. However, these estimates lack important information, such as crop height. By combining six maps with detailed 3D crop surface models, CSMs, advanced methods can be used for predicting crop yields at various growth stages. An unmanned aircraft system, UAS, was used to capture standard RGB imagery datasets for corn grain yield prediction at three early-dash, mid-dash, and late-season growth stages. Imagery was processed into orthophotos for crop slash non-crop classification and 3D CSMs, for crop height determination at different spatial resolutions. Three linear regression models were tested on their prediction ability, using site-specific, I, unclassified mean heights, 2, crop-classified mean heights, and, 3, a combination of crop-classified mean heights with corresponding crop coverages. The models showed determination coefficients, R2, of up to 0.74, while model, 3, performed. This article was authored by Jacob Giebel, Johanna Link, and Wilhelm Kloppian. We are article.tv, links in the description below.