The cardioprotection afforded by remote ischaemic conditioning (RIC) is mediated via a complex mechanism involving sensory afferent nerves, the vagus nerve, and release of a humoral blood-borne factor. However, it is unknown whether release of the protective factor depends on vagal activation or occurs independently. This study aimed to evaluate the co-dependence of the neural and humoral pathways of RIC, focussing on the vagus nerve and intrinsic cardiac ganglia. In the first study, anesthetised rats received bilateral cervical vagotomy or sham-surgery immediately prior to RIC (4 × 5 min limb ischemia–reperfusion) or sham-RIC. Venous blood plasma was dialysed across a 12–14 kDa membrane and dialysate perfused through a naïve-isolated rat heart prior to 35-min left anterior descending ischemia and 60-min reperfusion. In the second study, anesthetised rats received RIC (4 × 5-min limb ischemia–reperfusion) or control (sham-RIC). Dialysate was prepared and perfused through a naïve-isolated rat heart in the presence of the ganglionic blocker hexamethonium or muscarinic antagonist atropine, prior to ischemia–reperfusion as above. Dialysate collected from RIC-treated rats reduced infarct size in naïve rat hearts from 40.7 ± 6.3 to 23.7 ± 3.1 %, p < 0.05. Following bilateral cervical vagotomy, the protection of RIC dialysate was abrogated (42.2 ± 3.2 %, p < 0.05 vs RIC dialysate). In the second study, the administration of 50-μM hexamethonium (45.8 ± 2.5 %) or 100-nM atropine (36.5 ± 3.4 %) abrogated the dialysate-mediated protection. Release of a protective factor following RIC is dependent on prior activation of the vagus nerve. In addition, this factor appears to induce cardioprotection via recruitment of intrinsic cardiac ganglia.Electronic supplementary materialThe online version of this article (doi:10.1007/s00395-016-0568-z) contains supplementary material, which is available to authorized users.