Background: Cerebrovascular reactivity has been identified as an important contributor to secondary injury following moderate to severe traumatic brain injury (TBI). “Gold-standard” intracranial pressure (ICP) based indies of cerebrovascular reactivity are limited by their invasive nature poor spatial resolution. Near infrared spectroscopy (NIRS) based indices of cerebrovascular reactivity are minimally invasive and have improved spatial resolution. In this study, classical machine-learning algorithms are leveraged to better characterize the relationship between these indices. Methods: High-resolution physiologic data was collected in a cohort of adult moderate to severe TBI patients. From this data both ICP and NIRS based indices of cerebrovascular reactivity were derived. Utilizing Agglomerative Hierarchical Clustering (AHC) and Principal Component Analysis, the relationship between these indices in higher dimensional physiologic space was examined. Results: A total of 83 patients with 314,395 minutes of unique and complete physiologic data was obtained. Through AHC and PCA there was higher order clustering between NIRS and ICP based indices, separate from other physiologic parameters. Conclusions: NIRS and ICP based indices of cerebrovascular reactivity relate to one another in higher dimensional physiologic space. NIRS based indices of cerebrovascular reactivity may be a viable alternative to ICP based indices.