To assess the safety and accuracy of the Camino intraparenchymal sensor, we prospectively evaluated hemorrhagic complications, zero-drift, infection, and system malfunction in 163 patients monitored after a severe head injury. Mean duration of intracranial pressure (ICP) monitoring was 5 +/- 2.2 days (range: 12 h to 11 days). Of the 141 patients with a control CT scan, four showed a 1-2-cc collection of blood at the catheter's end. When removed, the sensors underread the true ICP value (negative zero-drift) in 80 of the 126 sensors evaluated (63.5%). Fourteen sensors showed no zero-drift, and 32 sensors overread the true ICP value (positive zero-drift) (median: -1 mm Hg; interquartile range: -4 to +1 mm Hg). No significant relationship was found between zero-drift, the surgeon who implanted the sensor, intracranial hypertension, or duration of ICP monitoring. No clinical infections could be attributed to the devices. Sixteen patients (9.8%) required more than one ICP sensor due to malfunctioning of the system. In conclusion, continuous ICP monitoring using the Camino intraparenchymal sensor has a low complication rate. However, this sensor may underread the real ICP values in a high number of patients. The lack of correlation between duration of ICP monitoring and zero-drift suggests that, contrary to the recommendations of other reports, the intraparenchymatous Camino sensor can provide reliable readings after the fifth day of use.