Résumé : Near-infrared spectroscopy (NIRS) permits non-invasive evaluation of tissue oxygen saturation (StO2). A vascular occlusion test (VOT) produces transient controlled ischemia similar to that used in ischemic preconditioning. We hypothesized that we could evaluate local responses to ischemic preconditioning by performing repeated VOTs and observing the changes in different NIRS VOT-derived variables. In healthy volunteers (n=20), four VOTs were performed at 30-min intervals on one day and, in a second group (n=21), two VOTs with time intervals of 5, 15 or 30min were performed on 3 separate days. Two cohorts of patients, one with circulatory shock (n=23) and a hemodynamically stable group (n=20), were also studied, repeating the VOT twice with a 5-min interval. In the 1-day volunteers, there was a median decrease of 15 (6-21)% in the Desc slope (StO2 decrease during VOT) after the second VOT, but no significant change in the Asc slope (StO2 increase after VOT). In the 3-day volunteers, the Desc slope also decreased, regardless of the time interval between VOTs. There was no overall decrease in the Desc slope in either patient cohort with repeated VOTs but there was marked individual patient variability. Patients in whom the Desc slope decreased had less organ dysfunction at admission, required less norepinephrine (0.00 vs 0.08mcg/kg/min, p=0.02), less frequently had sepsis (12 vs 50%, p=0.02) and had a lower mortality (6 vs 39%, p=0.03) compared to those in whom it did not decrease. Repeated NIRS VOT can non-invasively assess the local effects of ischemic preconditioning in the muscle.