Geometric morphometrics has been integral to understanding how form and function have evolved. However, traditional landmarking methods are restrictive, which can be a challenge to capture complex three-dimensional shapes. Automated landmarking can alleviate some of these challenges by placing pseudo-landmarks across an entire surface to capture variation in bones with few homologous regions. The neotropical loricariid catfishes (suckermouth armored catfishes or plecos) have ventrally placed oral jaws that are used to scrape food particles from surfaces. The upper jaw consists of a highly mobile premaxilla that is controlled by the maxillary motion via a unique branch of the adductor mandibulae that rotates the upper jaw around a process of the mesethmoid. The lower jaw is comprised of medially separated mandibles that rotate around the long axis within a shallow socket at the anteroventral articulation of the quadrate, which can be operated separately. Across 1006 species of the Loricariidae, jaw shape is highly variable and structurally complex, ranging from short jaws with less than five teeth to long jaws with over 200 teeth. The resulting differences in key characteristics of the jaws compound the difficulty of assessing homologous locations on the jaws for traditional geometric morphometrics. To test the effectiveness of automated landmarking, we isolated oral jaws from 36 species distributed across the phylogeny of armored catfishes from CT scans obtained from the MorphoSource database. Morphospace and phylomorphospace comparisons suggest that automated landmarking methods are better able to group like shapes compared to traditional methods. Nonetheless, automated methods require similar regions to automatically align models and are more susceptible to flipping shapes, which may cause issues in further shape analyses.