Four hundred and sixty-seven cases with implantation of an artificial pacemaker were studied. The postoperative survival rate was 63% for 15 years. Seventy-two percent of type III patients of the sick sinus syndrome were free from postoperative thromboembolism and the lowest of the three types of the sick sinus syndrome. Comparing postoperative physical activity, cardiothoracic ratio and exercise tolerance time, physiological pacing was superior to ventricular pacing in hemodynamic effects and clinical symptoms. In a hundred cases of physiological pacing, complications and problems of physiological pacing were discussed. Atrial sensing failure and over-sensing were observed in seven and two cases respectively. A low amplitude of atrial potential and use of unipolar atrial leads were considered to be the main causes of these complications. Bipolar lead should be used as the atrial lead to avoid such complications, because the atrial potential by bipolar leads is not less than that by unipolar leads. Atrial sensing may be more sensitive without electromagnetic interference. The fixed A-V delay time whenever the atrium is sensed or paced, often results in a ventricular fusion beat and hemodynamic change on every beat, according to the interval of atrial and ventricular contractions. The A-V delay time should be changed in accordance with atrial sensing or pacing.