Sepsis, the systemic inflammatory response to infection, is frequently complicated by brain dysfunction, which may involve disturbances in dynamic cerebral autoregulation (dCA) (Berg et al. 2011). We have previously established lipopolysaccharide (LPS) infusion as a human-experimental model of systemic inflammation that mimics the early stages of sepsis (Taudorf et al. 2007). We hypothesised that LPS infusion would impair dCA in healthy volunteers, and that patients suffering from full-scale sepsis would exhibit a similar pattern. Nine healthy male volunteers aged 23 (mean, SD 2) years, and 16 septic patients aged 59 (mean, SD 12; one female) years were enrolled in the study. Volunteers underwent a four-hour intravenous infusion of LPS (total dose 2 ng/kg). A 20-minute recording of invasive mean arterial blood pressure (MAP) and middle cerebral artery blood flow velocity (MCAv) measured by transcranial Doppler ultrasonography was performed at 1 kHz for evaluating dCA prior to and immediately after LPS infusion in volunteers, and within 72 hours of the sepsis diagnosis in patients. Transfer function gain, phase and coherence function between MAP and MCAv were calculated in the low frequency range (0.07-0.20 Hz) (Zhang et al. 1998). A decrease in coherence and gain, and an increase in phase were interpreted to reflect improved dCA and vice versa. Data are reported as median (interquartile range). LPS infusion induced a systemic inflammatory response with leukocytosis, fever, flu-like symptoms (all P<0.001), and increases in arterial cytokine levels (Figure). This was accompanied by increased heart rate, a decrease in MAP (both P<0.05), as well as a hyperventilatiory response (Baseline PaCO2: 5.4 [5.3-5.6] kPa; LPS PaCO2: 4.8 [4.6.5.0] kPa; P < 0.001). A decrease in gain and coherence, and an increase in phase were observed after LPS; in patients, both coherence and phase were lower than in healthy volunteers (Table). The cerebral haemodynamic aberrations may differ between the early and advanced stages of sepsis; contrary to our working hypothesis, the present findings indicate that dAR is improved after LPS-infusion, which is likely explained by hyperventilation, where the picure is less clear in patients.