We have performed an electrophysiologic study (EPS) in 18 subjects (mean age 21 +/- 17 years) with normal sinus node function (group A) and in 15 patients (mean age 43 +/- 26 years) with sinus node dysfunction (group B). Three extrastimulus methods (both Strauss and Kirkorian methods and our modification of the extrastimulus method) were carried out in 31 patients of both groups and sinoatrial conduction times (SACTSTR, SACTKIR,SACTMOD) could be estimated in 27 of them. SACT by the Narula method (SACTNAR) was assessed in 25 patients of both groups. Sinus node electrograms (SNEs) were attempted in 22 patients and were obtained in 17 of them. Among the indirect methods, by comparison with SACTSTR in group A the best correlation was for SACTMOD (r = 0.81) and the weakest for SACTKIR (r = 0.67); there was no significant difference between SACTs estimated by Strauss and other methods. In group B the weakest correlation was also for SACTKIR (r = 0.68), whereas the correlations for both SACTMOD and SACTNAR were similar (r = 0.72 and r = 0.74, respectively); in this group SACT was significantly underestimated only by the Kirkorian method (P less than 0.01; sensitivity of 45%). By our method sensitivity was 82%, similar to both the Strauss and Narula methods (92 and 92%, respectively). By comparison of indirect methods with the direct measurement of SACT we have found similar good correlations for the Strauss, our own, and the Narula methods in group A (r = 0.81, r = 0.75, r = 0.79, respectively) and in group B (r = 0.81, r = 0.86, and r = 0.72, respectively); here too the weakest correlation was, in both groups, for the Kirkorian method (r = 0.67, and r = 0.68, respectively). The SACT in group B was the more significantly underestimated (P less than 0.001). Significant underestimation of SACT occurred also by our method and the Strauss method (P less than 0.01, and P less than 0.001, respectively), whereas the Narula method did not significantly underestimate the directly measured SACT. From the SNEs after atrial pacing we have shown the limitation of the Kirkorian method and observed the conspicuous beat-to-beat variability of SACTdirect' as the possible explanation of chaotic pattern of return cycles by the extrastimulus methods in some patients of group B.(ABSTRACT TRUNCATED AT 400 WORDS)