When pacing trains with a constant cycle length (CL) but increasing number of beats are introduced during a macroreentrant atrial tachycardia (MAT), the postpacing interval (PPI) is expected to increase if entrainment does not occur but could be stable if entrainment occurs. We tested the ability of PPI analysis to detect entrainment.Synchronized pacing trains with increasing number of beats (1-20) were delivered from the coronary sinus (CS) and lateral right atrium (RA) at a CL 20 ms shorter than the MAT CL. Pacing trains were grouped in pairs differing by one-paced beat, and the ∆PPI measured.In an initial prospective cohort of 21 patients (48% had left atrial flutter) the mean ∆PPI was 21.3 ± 5.6 ms for pairs of pacing trains in which neither entrained the MAT and 2.8 ± 1.4 ms for those in which both entrained the MAT (P < .0001). Results were similar for common vs atypical flutter, PPI-TCL ≤30 ms vs PPI-TCL >30 ms, presence vs absence of antiarrhythmic drugs and faster vs slower MAT. When an index pacing train was compared to one with two-paced beats less, a PPI difference of <19 ms identified entrainment with 95% sensitivity and 98% specificity. In a validation cohort of 16 patients, this cut-off value resulted in sensitivity and specificity of 90% and 94%.A relatively constant ∆PPI in response to overdrive pacing with identical CL but different number of beats allows accurate discrimination between trains that entrained vs those which did not entrain a MAT.