The diagnostic value of programmed pacing in the investigation of tachycardia (greater than 5 premature complexes) was studied in 155 patients admitted for evaluation of dizzy attacks and/or tachycardia by determining a relationship between the induction of tachycardia by endocavitary pacing and the presence of spontaneous attacks on ECG and/or 24 hour Holter monitoring. Right atrial and ventricular programmed pacing comprised: an extrastimulus during sinus rhythm (method I), during paced rhythm (method II), 2 extrastimuli during sinus rhythm (method III) and paced rhythm (method IV). The protocol was applied in 20 cases of spontaneous atrial tachycardia (AT) and 40 patients without tachycardia, and in 20 cases of spontaneous sustained ventricular tachycardia (VTS) (Group A), 15 cases of non-sustained ventricular tachycardia (VTNS) (Group B), 20 cases of ventricular doublets or triplets on Holter monitoring (Group C) and 40 patients without ventricular arrhythmias. The following results were obtained: At atrial level, method I was associated with a 75% sensitivity and a 62.5 p. 100 specificity when the triggering of atrial echos was considered. It was difficult to induce AT with methods I, II and III (sensitivity 15, 20 and 45 p. 100 - but they were very specific (greater than 90 p. 100). The induction of echos with methods II, III and IV was very sensitive but not specific and could not be retained as a pathological criterion. Using method IV, only the triggering of sustained AT could be considered to have a good specificity (90 p. 100), but sensitivity remained low (30 p. 100). At ventricular level, more aggressive methods were needed to induce an arrhythmia.(ABSTRACT TRUNCATED AT 250 WORDS)