With the intravenous 133Xe technique we measured cerebral blood flow (CBF) in eight healthy subjects during normal subcutaneous temperatures and during extracranial cooling. This gave rise to the possibility of evaluating the contribution of the extracerebral blood flow to the calculation of CBF values. With a two-compartmental analysis of the wash-out curves during cooling there was a significant reduction of the CBF indices f1, representing mainly fast blood flow in the grey matter and f2, representing blood flow in the slowly perfused white matter and extracerebral structures. The reduction of f1 was due to the 'slippage' phenomenon:calculation of f1 was affected by a reduction in f2 due to a considerably reduced extracerebral blood flow. The initial slope index (ISI) calculated from 30 to 90 s of the first part of the presumed mono-exponential 133Xe wash-out curve was not affected by slippage as the ISI remained unchanged in spite of reduced extracerebral blood flow. It is concluded that CBF was unaffected by extracranial cooling. Extracranial cooling can be used to reduce the extracerebral blood flow contribution to the calculated CBF values.