Transorbital ultrasonographic measurement of optic nerve sheath diameter (ONSD) is an important technique for bedside assessment of raised intracranial pressure (ICP). However, developing competency for this clinical skill requires practice scans on both normal subjects and patients with raised ICP. The aim of this study is to develop a high-fidelity training model capable of measuring dynamic changes in ONSD and to test the reliability and reproducibility of the model at different simulated ICP values. We designed and developed a high-fidelity training model for dynamic ONSD measurement using a hemispherical table tennis ball, mounted on a 3.0-mm pediatric microcuffed endotracheal tube (ETT). Two independent investigators then performed a randomized blinded study to assess the reliability and reproducibility of the model. A total of 30 ONSD measurements (10 measurements each for 3 ETT cuff volumes of 0.1, 0.2, and 0.3 mL, simulating an ONSD of a normal, borderline, and raised ICP, respectively) were performed by each investigator. Intraclass correlation coefficients and Bland-Altman plots were calculated to analyze the level of agreement between the investigators. Our model was able to provide dynamic changes in ONSD secondary to ETT cuff volume changes. Small increments of 0.1 mL cuff volume changes produced immediate changes in ONSD that are similar to those observed in patients. The median interobserver difference in ONSD was 0.3 mm (interquartile range, 0. to 0.4 mm). Intraclass correlation coefficient was 0.89, 0.89, and 0.90 for 0.1, 0.2, and 0.3 mL ETT cuff volumes, respectively. We have developed a clinically relevant model capable of simulating changes in ONSD in patients with normal and raised ICP. This model could be a valuable training tool to gain scanning experience in optic nerve ultrasonography, and improve operators' technical abilities.