A screening examination including an electrocardiogram (limb leads only) coded by the Minnesota Code, using rigorous quality control was done on 18 403 male civil servants aged 40 to 64. The association of the findings with coronary heart disease has been tested in relation to age trends, symptomatic history, and coronary heart disease mortality rates in the ensuing five years. The results were positive for Q waves, left axis deviation, ST depression, and T wave changes (including minor T wave items as an isolated finding), ventricular conduction defects, and atrial fibrillation; but they were generally unimpressive for increased R amplitude and for lengthening or shortening of the PR interval, QT interval duration, premature beats, and extremes of heart rate. The prognosis of specific electrocardiographic findings discovered at screening is quite different from when they arise in clinical practice. Among the 6 per cent of men in this study with patterns suggesting ischaemia, the subsequent coronary heart disease mortality was little more than 1 per cent per year; and among those who were symptom free it was even less.