Background Single premature beats were introduced in the reentrant circuit during stable atrial flutter in the canine sterile pericarditis model to test the hypotheses that (1) despite the fact that the reentrant circuit is functionally determined, there is a fully excitable gap; (2) the excitable gap in the reentrant circuit is not uniform; and (3) inhomogeneities of conduction in the reentrant circuit explain the effects of premature beats.Methods and Results A multiplexing system was used to record 190 unipolar electrograms from the right atrial free wall during 18 atrial flutter episodes in 9 dogs. In all 18 episodes, premature stimuli captured the atrial flutter reentrant circuit. At the longest coupling intervals, the return cycle at the site closest to the pacing site did not prolong. As the coupling interval of the premature stimulus decreased, the return cycle then progressively increased, associated with changes in conduction in the reentrant circuit that were not uniform. The result was that coupling intervals associated with introduction of the premature beat also were not constant. The mean duration of the total (ie, fully plus partially) excitable gap was 12+-4 ms in areas of slow conduction, and it was always shorter than the total excitable gap in other areas (22+-6 ms, P<.001). The mean duration of the fully excitable gap based on analysis of the return cycle was 4+-1 ms in the reentrant circuit. In 13 of 18 atrial flutter episodes, a premature stimulus terminated atrial flutter by causing block of the orthodromic wave front of the premature beat in an area of slow conduction. The mean coupling interval that caused orthodromic block was 113+-5 ms (recorded at the site just proximal to the area of block), and it was always longer than the delivered stimulus coupling interval at the pacing site (96+-8 ms, P<.001).Conclusions We conclude that in this functionally determined atrial flutter reentrant circuit in the canine sterile pericarditis model, (1) a fully excitable gap is present in at least part of the reentrant circuit; (2) the duration of the excitable gap in the reentrant circuit is shortest in areas of slow conduction; and (3) when a premature beat encounters the partially excitable gap of the reentrant circuit, it results in changes in conduction such that the coupling intervals are not uniform throughout in the reentrant circuit. (Circulation. 1994;90:1997-2014.)