Gait analysis of transgenic mice and rats modeling human diseases often suffers from the condition that those models exhibit genotype-driven differences in body size, weight, and length. Thus, we hypothesized that scaling by the silhouette length improves the reliability of gait analysis allowing normalization for individual body size differences. Here, we computed video-derived silhouette length and area parameters from a standard markerless gait analysis system using image-processing techniques. By using length-and area-derived data along with body weight and age, we systematically scaled individual gait parameters. We compared these different scaling approaches and report here that normalization for silhouette length improves the validity and reliability of gait analysis in general. The application of this silhouette length scaling to transgenic Huntington disease mice and Parkinson´s disease rats identifies the remaining differences reflecting more reliable, body length-independent motor functional differences. Overall, this emphasizes the need for silhouette-length-based intra-assay scaling as an improved standard approach in rodent gait analysis.