Baroreceptor responses are gated during inspiration and efferent vagal supply to the heart is modulated by respiratory activity. Such variations in vagal tonus underlies cardio-respiratory coupling in mammals, but it has been suggested that such central control is absent in non-mammalian vertebrates (Porges, 1995, 2003). We examined Antarctic (environmental temperature -1.8 to 0°C), sub-Antarctic and temperate fishes (8 to 12°C) of similar genotype and ecotype, but which face vastly different selection pressures related to oxygen supply due to their different habitat temperatures. If cardio-respiratory coupling were present to optimise oxygen uptake, then it may be expected that a difference would be observed between these thermal groups. Fish were anaesthetised (MS222, Sigma; 0.5 g l^-1) and e.c.g. recording electrodes inserted close to the pericardium (0.02mm diameter, multi-strand Teflon coated stainless steel wire). Fish were allowed to recover until resting vagal tonus was re-established (determined by atropine-sensitivity in heart rate (fH) and development of significant heart rate variability (HRV). All fish were humanely killed at the end of the experiment. The Fourier (time-to-frequency) transform was applied to the e.c.g. tachogram, where sampling frequency for an individual fish was determined by its own fH. The intrinsic fH for all the fish species studied was ~ 25% greater than ventilation rate (fV), but vagal activity produced a resting fH that was synchronised with fV in a progressive manner. Coefficient of variation in the R-R tachogram was similar between groups. Power spectral statistics showed that episodes of relative bradycardia occurred in a cyclical manner, occurring every 2-4 heart beats in temperate species but >4 heart beats in Antarctic species. The ratio of LF/HF components provides an index of the dominant oscillations responsible for overall HRV. This ratio was always 1 in Antarctic fish. These oscillatory components were controlled centrally by vagal tonus, as bilateral cardiac vagotomy abolished all peaks. Although the modulating effects of ventilation were acting in a similar manner for all fish to coordinate blood and water flow at the gills, the high oxygen content and/or the low metabolic rate as consequence of living in sub-zero water, results in a relaxed selection pressure for cardio-respiratory coupling in Antarctic species. We conclude that vagally mediated control of fH is present in fish and operates around the less variable ventilatory cycle, similar to that controlling respiratory sinus arrhythmia in mammals.