Objective: The purpose of this study was to develop a novel pelvic model and determine the accuracy and the inter- and intraexaminer reliability of anterior superior iliac spine (ASIS) positional asymmetry assessment from both sides of the model by osteopathic predoctoral fellows and osteopathic physicians and to evaluate the effect of training. Methods: Five osteopathic predoctoral fellows and 5 osteopathic physicians assessed 13 settings of varied ASIS asymmetry of a novel pelvic model for superior/inferior positional asymmetry from both sides of the model in a random order. Assessment from the right and left sides of the model occurred on 2 separate days. Fellows were trained for a week and retested. Results: Average interexaminer reliability was greatest from the left side of the model for physicians and from the right side for fellows (physicians: κ = 0.46, fellows: κ = 0.37), whereas intraexaminer reliability was greatest from the right in both groups (physicians: κ = 0.49, fellows: κ = 0.52). Following training of fellows, interexaminer reliability remained highest from the right side of the model (right: κ = 0.48, left: κ = 0.36), whereas intraexaminer reliability was higher from the left side (right: κ = 0.53, left: κ = 0.59). Physicians and fellows before training were more accurate from the right side of the model (κ = 0.56 and κ = 0.52, respectively). Following training of fellows, accuracy increased from both sides of the model (right: κ = 0.59, left: κ = 0.53). Conclusions: A novel pelvic model was developed to allow assessment of accuracy and reliability of ASIS asymmetry assessment. Individually, physicians and fellows varied in accuracy and inter-/intraexaminer reliability. Further investigation is warranted to understand the clinical and educational application of these results. © 2010 National University of Health Sciences.