Abstract Background Regional oximetry is increasingly used to monitor post-extraction oxygen status of the brain during surgical procedures where hemodynamic fluctuations are expected. Particularly in cardiac surgery, clinicians employ an interventional algorithm to restore baseline regional oxygen saturation (rSO2) when a patient reaches a critical desaturation threshold. Evidence suggests that monitoring cardiac surgery patients and intervening to maintain rSO2 can improve postoperative outcomes; however, evidence generated with one manufacturer’s device may not be applicable to others. We hypothesized that regional oximeters from different manufacturers respond uniquely to changes in oxygen saturation in healthy volunteers. Methods Three devices were tested: INVOS™ 5100C (Medtronic), EQUANOX™ 7600 (Nonin), and FORE-SIGHT™ (CASMED) monitors. We divided ten healthy subjects into two cohorts wearing a single sensor each from INVOS and EQUANOX (n = 6), or INVOS and FORE-SIGHT (n = 4). We induced and reversed hypoxia by adjusting the fraction of inspired oxygen. We calculated the magnitude of absolute rSO2 change and rate of rSO2 change during desaturation and resaturation, and determined if and when each device reached a critical interventional rSO2 threshold during hypoxia. Results All devices responded to changes in oxygen directionally as expected. The median absolute rSO2 change and the rate of rSO2 change was significantly greater during desaturation and resaturation for INVOS compared with EQUANOX (P = 0.04). A similar but nonsignificant trend was observed for INVOS compared with FORE-SIGHT; our study was underpowered to definitively conclude there was no difference. A 10% relative decrease in rSO2 during desaturation was detected by all three devices across the ten subjects. INVOS met a 20% relative decrease threshold in all subjects of both cohorts, compared to 1 with EQUANOX and 2 with FORE-SIGHT. Neither EQUANOX nor FORE-SIGHT reached a 50% absolute rSO2 threshold compared with 4 and 3 subjects in each cohort with INVOS, respectively. Conclusions Significant differences exist between the devices in how they respond to changes in oxygen saturation in healthy volunteers. We suggest caution when applying evidence generated with one manufacturer’s device to all devices.