We investigated the influence of Forced Diffusion Ventilation (FDV), a special form of High Frequency Ventilation (HFV), on elevated intracranial pressure (ICP) in 5 dogs. Elevation of ICP was standardized by inflation of an epidural balloon. A typical finding with FDV is a reduced intrapleural pressure and therefore one could expect a better cerebrovenous drainage influencing ICP. Nevertheless, we found no changes in mean ICP under conditions of FDV compared with IPPV. Respirator-synchronous fluctuations of ICP, cisternal cerebrospinal fluid pressure and intrapleural pressure were drastically reduced with FDV. This phenomenon has been already reported by other groups as a typical effect of HFV with rates of 100/min. One can speculate, that this immediate impact of HFV on ICP-curves might be of some advantage in patients with critically reduced intracranial compliance requiring long-term artificial ventilation, because peaks and amplitudes of ICP are reduced. Our clinical experience with High Frequency Pulsation (HFP) includes 11 patients with severe brain trauma. In clinical routine this method of HFV is more facile to applicate than FDV, because there is no need of a special endotracheal tube and sufficient CO2-elimination is not strongly dependent on precise position of the tube. But HFP, as FDV, includes all advantages of respiratory systems, that are open against atmosphere (coughing and simultaneous breathing, without drastically increasing airway pressure, suction during respiration, etc.). However, we could find no special advantages or disadvantages in ICP-course during long-term application of HFP (up to 10 days). Because application of HFV is dependent on special technical equipment, we investigated in 6 patients the influence of respiratory frequency, tidal volume and inspiratory flow on ICP-fluctuations using conventional ventilators. ICP was recorded by a new, self constructed pneumatic epidural pressure sensor. Ventilator-related ICP-fluctuations were found to be markedly reduced at frequencies of 20/min and usually eliminated at 30/min. We found an exponential correlation between ICP-fluctuations and respiratory frequency and there was no correlation between tidal volume and ICP. Central venous pressure amplitudes were found to be in linear correlation with respiratory frequency and tidal volumes as well. The amplitude of respiratory ICP-fluctuations seems to be more dependent on duration of expiratory time. As our findings demonstrate, artificial ventilation without entilator-related fluctuations in ICP ("brain-protective" ventilation) may be performed by conventional volume-constant, time-cycled ventilators.(ABSTRACT TRUNCATED AT 400 WORDS)