A harmful effect of prolonged hyperventilation on outcome has been shown in comatose patients after severe head injury. The purpose of this study was to assess the acute effect of moderate hyperventilation for treatment of intracranial hypertension (ICP < 20 mmHg) on invasively measured brain tissue-PO2 (PtiO2), PCO2 (PtiCO2) and pH (tipH) in severely head injured patients. 15 severely head injured patients (GCS < or = 8) were prospectively studied. Intracranial pressure (ICP), mean arterial blood pressure (MABP), cerebral perfusion pressure (CPP), endtidal CO2 (ETCO2), PtiO2, PtiCO2 and tipH (Paratrend or Licox microsensors) were continuously recorded using multimodal monitoring. Following a baseline period of 15 minutes, patients were hyperventilated for 10 minutes. Arterial blood gas analysis was done before, during and after hyperventilation. At least three hyperventilation maneuvers were performed per patient. For statistical analysis the Friedman test was used. Hyperventilation (paCO2: 32.4 +/- 0.6 to 27.7 +/- 0.5 mmHg) significantly reduced ICP from 25.3 +/- 1.5 to 14.2 +/- 1.9 mmHg (p < 0.01). As a consequence, CPP increased by 9.6 +/- 3.4 mmHg to 76.8 +/- 3.2 mmHg. Brain tissue PCO2 decreased from 37.5 +/- 1.3 to 34.6 +/- 1.2 while tipH increased from 7.13 to 7.16. In all patients, hyperventilation led to a reduction of brain tissue PO2 (PtiO2/Licox: 24.6 +/- 1.4 to 21.9 +/- 1.7 mmHg, n.s.; PtiO2/Paratrend: 35.8 +/- 4.3 to 31.9 +/- 4.0 mmHg, n.s.). In one case hyperventilation even had to be stopped after 7 min because the drop in brain tissue PO2 below 10 mmHg signalized imminent hypoxia. As well known, hyperventilation improves CPP due to a reduction in ICP. However, this does not ameliorate cerebral oxygenation as demonstrated by the decrease in PtiO2. This underlines that hyperventilation should only be used with caution in the treatment of intracranial hypertension.