A series of 123 patients with valvular heart disease underwent treadmill stress testing (Bruce’s protocol). Neither mortality nor serious complications were observed. The incidence of exercise-induced ventricular ectopy was 35% and that of complex ventricular ectopy 19 %. Patients with pure left ventricular volume overload had a significantly higher incidence of exercise-induced ventricular ectopy than did patients with other valvular diseases. A significant relation was found between functional class (NYHA) and either duration of exertion or functional aerobic impairment. The NYHA classification was found to coincide in 62% of the instances with another based on the results obtained from stress testing. In patients with mitral stenosis a significant although slight correlation was found to exist between mitral valve area and duration of exercise. The inability to reach stage III of Bruce’s protocol was associated with a critical mitral stenosis. A hypotensive response to exercise was frequent both in mitral stenosis and aortic stenosis. But whereas in aortic stenosis it was characteristic of the most severe lesions, in mitral stenosis it bore no relation whatsoever to severity. Thus, treadmill stress testing in valvular heart disease is safe, can be helpful in quantifying functional disability and may provide a clue as to the severity of the lesion in specific subsets.