The place of balloon dilatation of the aortic valve in the treatment of calcific aortic stenosis is controversial. Thirty two patients (mean age 76) in whom valve replacement was contraindicated were followed up for three to 24 months (mean 8); 25 were in functional class III or IV according to the New York Heart Association's classification. Major complications of the procedure occurred in four patients. Echocardiography and Doppler studies were performed before operation and before discharge in 28 patients, and the area of the valve was measured again six to 50 (mean 23) weeks after operation in 11 patients. The peak to peak aortic pressure gradient fell from a mean of 65 (SD 24) to 46 (20) mm Hg, but the area of the aortic valve, measured by Doppler echocardiography, in 18 patients showed a modest but significant increase, from 0.61 (0.16) to 0.74 (0.23) cm2. One month after dilatation, 29 patients were alive, of whom 17 had improved symptoms. Only two had lasting clinical benefit. Sixteen patients died, 12 of a cardiac cause. The estimated one year survival rate was 49%. Six patients underwent or required valve replacement because of persisting symptoms. In view of its limited long term efficacy balloon dilatation of the aortic valve should be used only for patients with severe symptoms whose life expectancy is limited by other disease or who are considered to be unsuitable for valve replacement. It may have a role in improving the condition of patients who present with cardiogenic shock or pulmonary oedema before valve replacement is undertaken.