Four automated blood pressure machines (Arteriosonde 1225, Bosomat 11D, Elag Koln BE 237R and Omron HEM 3) have been compared with each other and with the Hawksley Random Zero Sphygmomanometer in 18 subjects with a wide range of BP. Each form of measurement provided a high degree of internal consistency. The Arteriosonde 1225 consistently recorded systolic and diastolic pressures which were below the values obtained with the Random Zero instrument (P=0.001). The other 3 automated machines recorded BPs which did not differ significantly from the Random Zero or from each other. Heart rate readings did not differ significantly for any of the machines compared to manual recording. The disparity attributable to the Arterisonde could result from its being the only form of measurement to use an ultrasonic rather than a microphonic recording technique. However, the cheaper and more portable automated machines produce results which closely reflect those obtained by using a mercury sphygmomanometer.