This study assessed the accuracy of oscillotonometric blood pressure measurement in critically ill patients during transfer. Direct intra-arterial pressure measurements were taken in 44 transported patients as a 'gold standard' and compared with readings from four portable automatic oscillotonometers--the Dinamap 8100, Lifestat 100, Propaq 102 and Takeda UA711. All under-read systolic pressure (by 13%, 21%, 19% and 13% respectively) and over-read diastolic pressure (by 15%, 5%, 27% and 15% respectively) in comparison to direct pressure measurement. The limits of agreement for 95% of comparisons were broad. The systematic difference between direct and oscillotonometric measurements and the variability between and within patients show that direct oscillotonometric measurements are not interchangeable in these patients. Portable automatic oscillotonometers should not be substituted for direct monitoring in such circumstances.