It has been suggested that respiratory sinus arrhythmia (RSA) contributes to the optimisation of pulmonary gas exchange efficiency by clustering heartbeats in the inspiratory phase of the respiratory cycle (Hayano’s hypothesis). This hypothesis is supported by animal and human studies, which show a significant correlation between RSA magnitude and indices of gas exchange efficiency (1, 2). However, recent studies indicate that RSA may play a limited role in clustering heartbeats (3, 4) and that changes in the proportion of heartbeats in inspiration (HB_insp) are more closely related to changes in the inspiratory period to breath period ratio (IE/II ratio) than RSA magnitude per se. In this study we sought to examine the pattern of heartbeat distribution throughout the respiratory cycle across a range of RSA magnitudes likely to be observed physiologically in man. In 12 healthy male volunteers (aged 20-25) we recorded ECG, respiratory flow and continuous BP measurements in the supine position and modified RSA magnitude by fixed paced breathing at 6, 9 and 12 breaths per minute. RSA pattern and magnitude were obtained by cubic spline interpolation of cardiac cycle intervals as a function of the respiratory cycle. One-way repeated measures ANOVA showed that reductions in breathing frequency resulted in a significant increase in RSA magnitude (p < 0.01), but this was not associated with significant changes in the proportion of heartbeats in inspiration, HB_insp (p = 0.33). Although IE/II ratio did not differ significantly with breathing frequency (p = 0.069), linear regression analysis showed a strong relationship between HB_insp and changes in IE/II ratio (r = 0.85, p < 0.01) but not with changes in RSA magnitude (r = 0.19, p = 0.55). These results suggest that, contrary to the common view, RSA magnitude does not cause significant heartbeat clustering into inspiration in humans. The mechanism behind associations between RSA and indices of gas exchange efficiency and the underlying function of RSA remain unclear.