This article provides a comprehensive guide for the use of electrocardiographic (ECG) assessments in clinical trials, for non anti-arrhythmic products, drawing together scientific information and current and proposed regulatory requirements and guidelines. Standard definitions of normal values of ECG intervals, ECG diagnostic criteria for myocardial infarction and left ventricular hypertrophy, and the frequency of expected ECG abnormalities are presented. A focus on the problem of drugs that affect the QTc interval including the regulatory concern and associated factors are reviewed. The high spontaneous variability in QTc duration requires careful attention to the planning and interpretation of ECG data if a true QTc effect of a new drug is to be detected adequately. Too much additional extrinsic variability may compromise the ability to detect a signal amongst the noise. The sources of extrinsic variability that must be considered in clinical research include: ECG acquisition methods; adequacy of ECG collection, especially at baseline and at steady state or peak drug concentration; sample size; presence of control groups; ECG measurement accuracy; and proper use of formulae for correcting QT to QTc. There are several compelling reasons to utilize centralized ECG analysis as the primary method of collation and analysis of 12-lead ECGs in clinical studies. While ECG effects are initially well defined in phase I studies, confirmation in phases II and III is still required since safety in the target population with co-morbidities and concomitant medications must be confirmed. All ECGs should be statistically analysed and reported according to pre-defined standards. These should be developed according to the specific project requirements, the known or suspected properties of the test compound and the intended treatment regimen. A standard analysis plan is proposed.