Purpose. To determine if a noninvasive brain acoustic monitor can differentiate acoustic responses from “normal patients” and ischemic from hemorrhagic stroke patients. Methods. A laptopsized passive acoustic monitoring system acquires arterial-pressure-generated signals during a 15-second monitoring session from sensors placed at the radial artery and on the fore-head. The arterial pulse waveform from the head is compared with that of the arterial waveform to generate the time-domain signal comparison. Frequency domain signals from each area are also compared. The study involved patients with diagnosis of first stroke who could be monitored within 12 hours of symptom onset and normal subjects who provided informed consent. Individuals with history of brain injury, stroke, or other brain disease were excluded. Results. Twelve normal subjects and 6 ischemic stroke, 2 transient ischemic attack (TIA), and 3 hemorrhagic stroke patients were monitored. Frequency response analysis identified uniform frequency responses in normal subjects. The signal in ischemic stroke patients was characterized by a divergence of the radial and cranial frequency response between 10 and 50 Hz of 10 dB or greater. In intracerebral hemorrhage patients, a divergence was seen below 10 Hz but not in the band above 10 Hz. TIA patients were monitored after symptom resolution and showed a divergence <10 dB in both bands, similar to normal subjects. Conclusions. In a pilot study using a noninvasive monitor, the authors detected a potential to differentiate between normal subjects and those with cerebral ischemia or hemorrhage.