The adequate management of cerebral perfusion pressure (CPP) continues to be a controversial issue in head-injured patients. The purpose of our study was to test two hypotheses. The first was that in patients with a CPP below 70 mm Hg, oxygen delivery is compromised and that therefore signs of tissue hypoxia would be reflected in low PtiO2 measurements. The second hypothesis was that manipulating mean arterial blood pressure to increase CPP improves oxygen delivery, particularly in patients with a CPP below 70 mm Hg. Twenty-five moderately or severely head-injured patients were included in the study. In all of them PtiO2 was monitored in the non-injured hemisphere using the Licox system (GMS, Kiel-Mielkendorf, Germany). Arterial hypertension was induced with phenylephrine 29 times. To quantify the effect of increasing mean arterial blood pressure (MABP) on oxygen delivery to the brain, the PtiO2-BP index was calculated (PtiO2-BP index = delta PtiO2/delta MABP). In 16 tests (55%) baseline CPP was above or equal to 70 mm Hg and in the remaining 13 (45%) it was below 70 mm Hg. Mean increase in MABP after phenylephrine was 23.7 +/- 10.2 mm Hg. Mean PtiO2 was 29.5 +/- 14.7 mm Hg in patients with a basal CPP of below 70 mm Hg and 28.9 +/- 10.6 mm Hg in patients in the high CPP group. These differences being not statistically significant. The PtiO2-BP index was 0.29 +/- 0.23 in patients with a basal CPP of below 70 mm Hg and in patients with a CPP of above 70 mm Hg this index was 0.16 +/- 0.11 Hg. These differences were not statistically significant (Student's t-test, P = 0.09). In our study a low PtiO2 was not observed in patients with marginally low CPPs (48-70 mm Hg) and readings below 15 mm Hg were observed in cases with both normal or supranormal CPPs. We conclude that episodes of low PtiO2 could not be predicted on the basis of CPP alone. On the other hand, raising CPP did not increase oxygen availability in the majority of cases, even if the CPP was markedly improved.