The effect of physical training on an individual's ability to withstand an orthostatic stress is unclear. This study was undertaken to determine the effects on orthostatic tolerance in healthy volunteers of training at a level appropriate for unfit subjects and cardiorespiratory patients. In 11 asymptomatic, untrained subjects the following assessments were made: plasma volume by Evans Blue dye dilution and blood volume derived from haematocrit; carotid baroreceptor sensitivity from the slope of the regression of change in cardiac interval against pressure applied to a neck chamber; orthostatic tolerance as time to presyncope in a test of head-up tilting combined with lower body suction; exercise test relating heart rate to oxygen consumption. Subjects were then given a training schedule (5BX/XBX, Royal Canadian Air Force) involving 11-12 min of mixed exercises per day until an age/sex related 'target' was reached. Following training all subjects showed evidence of improved fitness, seen as decreases in heart rate at an oxygen uptake (VO2) of 1Σ5 l min-1 and in the elevation of the regression line between heart rate and VO2. All also had increases in plasma and blood volumes and decreases in baroreceptor sensitivity. Seven of the subjects showed increases in orthostatic tolerance. Improvement in orthostatic tolerance was related to a low initial tolerance, and was associated with increases in plasma volume and decreases in baroreceptor sensitivity. These results show that moderate exercise training increases orthostatic tolerance in subjects who do not already have a high initial tolerance and suggest that training may be of value in the management of untrained patients with attacks of syncope due to orthostatic intolerance.