Cardiac dysfunction is highly prevalent in dialysis patients in the developed world, and is a major cause of morbidity and mortality. The relative impact of pre-existing cardiac disease and dialysis/uremia on cardiovascular morbidity are not clear. We conducted a retrospective, cross-sectional analysis of cardiac function and mortality in 202 incident and prevalent dialysis patients over an 18-month period in a population with a low prevalence of cardiovascular disease at dialysis initiation. Systolic dysfunction was defined as an ejection fraction (EF) of <50%. Left ventricular hypertrophy (LVH) was determined by echocardiography or electrocardiogram. Clinical data was collected by chart review. Ninety-nine percent of patients were black, with a mean age of 41.7 +/- 10.1 years, and median follow up 28 months (range 1-216 months). Echocardiograms were available in 132 patients. Seventy-seven patients received hemodialysis, and 55 received peritoneal dialysis. Mean EF was 63.2 +/- 11.1. EF was not lower in patients with greater duration of dialysis, although LVH tended to increase (not statistically significant). In 39 patients who died during the study period, cardiac function was not different from survivors, and no patient died of ischemic heart disease or heart failure. In conclusion, in a population of patients with a low prevalence of cardiovascular disease at dialysis initiation, cardiac function appears preserved over time, and cardiac morbidity and mortality are low. This finding suggests that dialysis and uremia per se, in the absence of pre-existing cardiac disease, may not be major contributors to cardiovascular morbidity.