The value of a 24-hour ambulatory electrocardiogram (AmECG) in assessing the risk of cardiac death was studied in 122 stable-condition dialysis patients followed up from two to six years after monitoring. An abnormal AmECG was defined by second-degree or greater AV block or by Lown grade 3 or greater ventricular ectopy. The incidence of cardiac mortality or an abnormal AmECG was not influenced by hypokalemia or beta-blockers. Digitalis was associated with both an abnormal AmECG and a twofold increase in mortality whether or not the AmECG was normal. Cardiac mortality accounted for 26 of 33 deaths within the first year after the AmECG. In the absence of coronary artery disease, survival at six months was 100% in patients with normal AmECG vs 90% (abnormal AmECG). In the presence of coronary artery disease, survival at six months was 83% (normal AmECG) vs 54% (abnormal AmECG).