INTRODUCTION: Normal pressure hydrocephalus (NPH) commonly presents with a triad of gait disturbance, incontinence, and memory impairment. Shunt placement effectively treats symptoms; however, NPH is commonly misdiagnosed with below 20% of patients receiving a diagnosis. Given undiagnosed NPH elderly patients often present to the emergency room (ER) with falls, algorithmic identification of NPH using non-contrast CT (NCCT) may improve diagnosis rate. METHODS: NCCTs for 84 patients with NPH, 72 headache controls (HC), 97 Alzheimer’s Disease (AD) patients, and 61 post-traumatic encephalomalacia (PTE) patients were identified using Veterans Affairs Informatics and Computing Infrastructure. Image processing pipelines were developed to extract a novel feature capturing lateral ventricle eccentricity (MaxEccLV), a proxy-Splenial Angle (p-SA), the Evans indices (EI-x, y, z), Callosal angle (CA), third ventricle width (NMax-3VW), and CSF to brain volume ratio (CSF2BVR). T-tests were used to examine group differences and logistic regression to classify between NPH, AD, PTE, and HC. RESULTS: The AUC, sensitivity, and specificity for NPH versus AD+HC+PTE was 0.94, 83.0%, and 87.0% respectively. For NPH versus HC, it was 0.98, 98.8%, and 98.6% respectively. For NPH versus AD, it was 0.90, 84.5%, and 82.5% respectively. For NPH versus PTE, it was 0.95, 91.7%, and 86.9% respectively. Significant differences were seen between NPH and HC, AD, and PTE in MaxEccLV and all other features. CONCLUSIONS: The computational pipeline proposed here can identify NPH from HCs and ventriculomegaly-causing conditions using NCCT with a high sensitivity and specificity. This may help capture undiagnosed NPH patients receiving a NCCT due to presentation to the ER as a fall.