OBJECTIVE To compare clinical and electrocardiographic characteristics of Nonsustained Ventricular Tachycardia (NSVT) and Idioventricular Accelerated Rhythm (IVAR). METHODS We studied 155 patients, 113 men and 42 women, with mean age 54 +/- 14 retrospectively, of these, 108 had NSVT and 47 IVAR. The arrhythmias were defined as follows: NSVT-more than 3 ventricular consecutive beats with an heart rate superior to 110 b/m and lasting less than 30 s.; IVAR-3 or more ventricular consecutive beats with an heart rate equal or superior to 50 and lower than 110 b/m, lasting less than 30 s. We evaluated clinical data (symptoms, functional class and anti-arrhythmic therapy), electrocardiographic data (rhythm, changes in conduction and repolarization) and ventricular function (with ECO, Radionuclide Angiography or Ventriculography). In the Holter recording (ECG-H), we analysed the presence of associated ventricular arrhythmias, their electrocardiographic characteristics (number of episodes, number of beats per episode, previous arrhythmia rate, morfology, regularity) and the relations of the arrhythmia with symptoms. RESULTS Analysis of underlying pathology showed in both groups, the importance of coronary artery disease (44.5% vs 40%) followed by valvular heart disease (24% vs 27.6%) and cardiomyopathy (22.2% vs 17%) respectively to NSVT and IVAR. Only in the NSVT group there were patients without cardiac pathology (3.6%). Comparing with one control group of our department, this distribution was substantially different (p less than 0.0001). All IVAR episodes were assympthomatic compared with 90% of NSVT. Ventricular premature beats were found in all NSVT patients and in 90% of IVAR patients, and were frequent (greater than 10/h) in 79% and 60%, couplets in 84% and 53% respectively (ns). The previous rate of the arrhythmia was 85.3 +/- 20 b/m in NSVT against 68.7 +/- 14 in IVAR (p less than 0.0001). We found left ventricular disfunction in 60% of NSVT patients and in 63.7% in IVAR patients, being serious in 35% and 39% respectively. The follow-up was of 18.5 months (1-72) and posterior evolution showed 14.8% and 17% of deaths with no relation to the arrhythmia, although in NSVT the number of complexes and episodes were related with the ventricular disfunction (p = 0.02 and p = 0.05). CONCLUSIONS Both arrhythmias appeared in patients with similar clinical and arrhythmic setting and identified a population with structural cardiopathy, bad function and poor outcome.