Perforated capsules were implanted aseptically into the brain parenchyma of dogs. The capsules became filled with fluid, and tissue accompanied by blood vessels grew within them. The internal fluid pressure of the intracerebral capsules (BC) was measured. Other perforated capsules were implanted into the subcutaneous tissue and muscle (denoted SC and MC), respectively. The fluid pressure within them was compared with that in BC. The following results were obtained:1) The average fluid pressure in BC was+57.9±8.3 (S. E. M.) mm H2O, and was entirely different from the-60.2±5.5 (S. E. M.) mmH2O in SC. A significant correlation was recognized between fluid pressure and protein content in the capsular fluids of SC and MC (P‹0.01) but not of BC (P›0.05).2) Bilateral occlusion of the external jugular veins induced a sudden increase in CSF (cerebrospinal fluid) and BC pressure. There was a return to control level on release.3) Increase in blood pressure caused only a slight elevation of BC and CSF pressure, with clear on-off effects in the case of carotid occlusion. Stimulation of the efferent vagus led to a decrease in blood pressure and to a small decrease in BC and CSF pressure, followed by marked fluctuation when the stimulus was removed. Low pressure changes in BC and CSF, concomitant with marked changes in blood pressure, are explained on the basis of an autoregulatory mechanism affecting the brain circulation.4) The half-time for pressure reequilibration, following injection into or removal of a small amount of fluid from the capsules, was measured. It was appreciably longer in BC than in SC.5) The evidence for a positive pressure in BC is discussed, and it is suggested that the BC pressure may reflect to some degree both the vascular and interstitial conditions of the brain.