The effect of changes in arm position and of arm support on blood-pressure readings was assessed in 240 hypertensive outpatients with the arm in various positions. Mean blood pressure was lowest with the arm supported at heart height and highest with the arm dependent, the difference averaging 8 mmHg for both systolic and diastolic pressure and being greater for men than for women. Arm support at heart height lowered blood pressure only slightly, when compared to that taken on the unsupported arm at heart height. These findings suggest that an increase in apparent blood pressure associated with arm dependency during the recording can spuriously increase the height of the blood pressure and the prevalence of hypertension. In the individual patient it may lead to unnecessary increase in antihypertensive dosage. Whenever standing blood pressure is quoted, arm position should be stated.