Near-Infrared Spectroscopy (NIRS) is an emerging medical countermeasure for rapid, field detection of hematomas caused by traumatic brain injury (TBI). Bench and animal tests to determine NIRS sensitivity and specificity are needed. However, current animal models involving non-invasively induced, localized neural damage are limited. We investigated an in vivo murine hematoma model in which cerebral hemorrhage was induced noninvasively by highintensity focused ultrasound (HIFU) with calibrated positioning and parameters. To characterize the morphology of induced hematomas, we used skull-intact histological evaluation. A multi-wavelength fiberoptic NIRS system with three source-detector separation distances was used to detect hematoma A 1.1 MHz transducer produced consistent small-tomedium hematoma localized to a single hemisphere, along with bruising of the scalp, with a low mortality rate. A 220 kHz transducer produced larger, more diffuse hematomas, with higher variability in size and a correspondingly higher mortality rate. No skin bruising or blood accumulation between the skin and skull was observed following injury application with the 220 kHz transducer. Histological analysis showed higher sensitivity for larger hematomas (>4x4 mm 2 ). NIRS optical density change after HIFU was able to detect all hematomas, with sensitivity dependent on wavelength and separation distance. While improvements in methods for validating cerebral blood distribution are needed, the HIFU hematoma model provided useful insights that will inform development of biologically relevant, performance test methods for cerebral NIRS systems.