In order to obtain direct evidence for tissue hypoxia in patients with sepsis oxygen, partial pressure was measured within skeletal muscle. Furthermore, serial intermittent and continuous measurements of skeletal muscle PO2 in patients with sepsis were used to find out whether skeletal muscle oxygenation may change in the course of sepsis and depends on the severity of sepsis. Prospective study. Intensive care unit of a university hospital. Intensive care patients (n = 98) with sepsis (group 1, n = 39; group 4, n = 28), limited infection (group 2, n = 16), and cardiogenic shock (group 3, n = 15). Pulmonary artery catheterization; standard antibiotic therapy and volume replacement. Skeletal muscle PO2 was determined by polarographic needle electrodes or catheter probes. In patients with sepsis (n = 67), no evidence for skeletal muscle hypoxia was obtained from the PO2 distribution within biceps muscle. Mean skeletal muscle PO2 was increased in patients with sepsis (group 1, 48 torr [6.4 kPa]) compared with patients with limited infection (group 2, 28 torr [3.7 kPa]), p < .001) and with patients with cardiogenic shock (group 3, 22 torr [2.9 kPa], p < .001). Serial measurements of the PO2 distribution during seven consecutive days in another 28 patients (group 4) with sepsis showed that a more severe degree of sepsis was associated with an increase of mean skeletal muscle (p < .001). These results were confirmed by continuous measurements of mean skeletal muscle PO2, using PO2 catheters. In patients with sepsis, oxygen transport to skeletal muscle was not critically reduced. Serial intermittent and continuous measurements of skeletal muscle PO2 showed that skeletal muscle PO2 increased in relation to the severity of the stage of sepsis. Our findings suggest that oxygen utilization within skeletal muscle decreased with deterioration of sepsis, thereby increasing skeletal muscle PO2.