The precise aetiology of sudden death in patients receiving neuroleptic medication is uncertain, but cardiac arrhythmias are a possible cause. We investigated the link between neuroleptic medication and electrocardiographic changes predictive of malignant cardiac arrhythmias. Electrocardiographs were performed on 111 patients receiving neuroleptic medication and on 42 unmedicated controls. Prolonged QTc intervals were more common in the patient sample, but QTc dispersion was not significantly increased. QTc interval prolongation was more likely in patients on doses above 2000 mg chlorpromazine equivalents daily (odds ratio 4.28, P < 0.02). Neuroleptic medication, especially at high doses, is associated with ECG changes that may herald more serious cardiac problems.