The results of 269 patients who had undergone valvular operations during and 18-month period (January 1, 1980, to June 30, 1981) were analysed. One hundred and seventy-four patients referred for surgery had undergone routine cardiac catheterisation elsewhere, and these, together with four electively catheterised patients, constituted a group representing the conventional approach. Ninety-one patients were assessed only by clinical and non-invasive methods, including echocardiography, and were subjected to surgery without catheterisation. The two groups were comparable in their major clinical characteristics including age, sex ratio, functional class, previous history of closed mitral valvotomy, aetiology of valve lesions, and type of operative procedure performed. There was no discordance between operative finding and preoperative assessment, except in two patients of the catheterised group. Hospital mortality was acceptably low and was comparable between the two groups. Cardiac catheterisation is, thus, no longer necessary in the majority of patients undergoing surgery for valvular heart diseases.