Respiratory sinus arrhythmia (RSA) is the oscillation of heart rate in phase with ventilation and represents the high frequency component of heart rate variability. RSA is prevalent in the young and the endurance-trained and its loss in heart failure is considered a prognostic indicator for sudden death. The recapitulation of RSA improves cardiac function in animal models of heart failure (O'Callaghan et al., 2020; Shanks et al., 2022). However, the precise mechanisms by which RSA affects cardiac function are unknown. We have previously shown that recapitulation of RSA through oscillatory pacing (OP), in which cycle length (CL) was set to 20% below and 20% above the average basic CL in a 2-short, 2-long stimulus train, entrained action potential duration (APD), Ca²⁺ transient amplitude (CaT) and sarcomere shortening and reduced the incidence of alternans in comparison to pacing at constant CL in guinea pig isolated ventricular myocytes (Carpenter et al., 2021; Carpenter et al., 2022). The OP entrainment of CaT and sarcomere shortening persisted in voltage clamp experiments with fixed APD (i.e. AP clamp), suggesting that APD is not an essential component of entrainment. Here, we examine the hypothesis that OP in isolated ventricular myocytes affects the stimulus-interval dependence of APD and sarcomere shortening through mechanisms involving the sarcoplasmic reticulum (SR). Animal procedures were approved by the University of Bristol Animal Welfare and Ethics Review Board and conducted according to UK legislation. Hearts were excised from adult male guinea pigs under general anaesthesia (140 mg/kg bodyweight euthatal, i.p.) and left ventricular myocytes isolated following enzymatic digestion. Isolated myocytes were superfused with Tyrode’s solution at 37 °C for perforated patch whole-cell current clamp recording and action potentials and sarcomere length measured simultaneously. Sample sizes are presented as n (number of cells) from N (number of animals) and nested hierarchical analysis applied using a generalized linear mixed model (SPSS, IBM). The limit of statistical confidence was P<0.05. Using a three-step S₁S₂S₃ protocol to examine the dependence of APD and sarcomere shortening on both preceding (CL_n-1) and anteceding (CL_n-2) stimulus intervals, both APD and sarcomere shortening were found to depend directly on the preceding cycle length (CL_n-1, P<0.001). In contrast, sarcomere shortening depended inversely on the anteceding CL (CL_n-2, P<0.001), while APD₉₀ did not depend significantly on CL_n-2 (n/N = 24/4). This is consistent with the entrainment by OP, the effect of CL_n-1 on sarcomere shortening being akin to post-extrasystolic potentiation (Yue et al., 1985). Following treatment of the cells with caffeine (50 µM) to sensitise ryanodine receptors, both APD and sarcomere shortening in response to the CL_n-1 stimulus were markedly suppressed, and the inverse dependence of sarcomere shortening on anteceding CL_n-2 was lost (P<0.001, n/N = 6/3). These data are consistent with an obligatory role for the SR in the entrainment of APD, CaT and sarcomere shortening and the suppression of alternans by the recapitulation of RSA. Further work is required on the effects of OP in intact hearts to determine the potential for a protective effect of RSA recapitulation against ventricular tachyarrhythmias.