This cross-sectional study aimed to provide standardized nomenclature for categorizing bladder shapes and find evidence that some bladder shapes are associated with high filling pressures and dysfunction in the pediatric spina bifida population. From July 2016 to July 2022, fluoroscopic images of 417 pediatric patients’ urinary systems at Children’s Hospital of Philadelphia were classified by a machine learning model and by urologists. The images were analyzed by a pre-trained DenseNet deep learning model via unsupervised cluster analysis to group the images into 5 clusters. 5 urologists independently determined 5 bladder shape classifications based on shape (round, oblong, and Christmas tree) and contour (smooth or trabeculated) using clinical expertise and literature review; 2 urologists manually grouped the same images based on these classifications. The primary outcome was bladder filling pressure and secondary outcomes included reflux, leak, and sphincter dysfunction. The model sorted bladders that appeared trabeculated into clusters that also had statistically significant higher filling pressures at most filling levels. Christmas tree shaped bladders were also associated with increased filling pressures and secondary outcomes. Despite these differences, general bladder dysfunction was not varied across groups. Finally, while machine learning and urologists were able to independently generate similar bladder classifications, no external validation of this approach was performed.