Vascular calcifications are frequent in chronic renal disease and are associated to significant cardiovascular morbidity and mortality. The long term predictive value of coronary artery calcifications detected by multi‐layer spiral computed tomography for major cardiovascular events was evaluated in non‐diabetic Caucasian patients on maintenance hemodialysis free of clinical cardiovascular disease. Two‐hundred and five patients on maintenance hemodialysis were enrolled into this observational, prospective cohort study. Patients underwent a single cardiac multi‐layer spiral computed tomography. Calcium load was quantified and patients grouped according to the Agatston score: group 1 (Agatston score: 0), group 2 (Agatston score 1‐400), group 3 (Agatston score 401‐1000) and group 4 (Agatston score >1000). Follow‐up was longer than seven years. Primary endpoint was death from a major cardiovascular event. Actuarial survival was calculated separately in the four groups with Kaplan‐Meier method. Patients who died from causes other than cardiovascular disease and transplanted patients were censored. The “log rank” test was employed to compare survival curves. One‐hundred two patients (49.7%) died for a major cardiovascular event during the follow‐up period. Seven‐year actuarial survival was more than 90% for groups 1 and 2, but failed to about 50% for group 3 and to <10% for group 4. Hence, Agatston score >400 predicts a significantly higher cardiovascular mortality compared with Agatston score <400 (p<0.0001); furthermore, serum Parathyroid hormone levels > 300 pg/l were associated to a lower survival (p < 0.05). Extended coronary artery calcifications detected by cardiac multi‐layer spiral computed tomography, strongly predicted long term cardiovascular mortality in nondiabetic Caucasian patients on maintenance hemodialysis. Moreover, it was not related to conventional indices of atherosclerosis, but to other non‐traditional risk factors, as serum Parathyroid hormone levels. A full cost‐benefit analysis is however necessary to justify a widespread use of cardiac multi‐layer spiral computed tomography in clinical practice.