We have previously shown that remote ischemic preconditioning (rIPC) by transient limb ischemia leads to the release of a circulating factor(s) that induces potent myocardial protection. Intra-arterial injection of adenosine into a limb also leads to cardioprotection, but the mechanism of its signal transduction is poorly understood. Eleven groups of rabbits received saline control or rIPC or adenosine administration with additional pretreatment with the nitric oxide (NO) synthase blocker N(G)-nitro-l-arginine methyl ester, the NO donor S-nitroso-N-acetylpenicillamine, its non-NO-donating derivative N-acetylpenicillamine, or femoral nerve section. Blood was then drawn from each animal, and the dialysate of the plasma was used to perfuse a naïve heart from an untreated donor. Infarct size was measured after 30 min of global ischemia and 120 min reperfusion. When compared with that of the control, mean infarct size was significantly smaller in groups treated with rIPC alone (P < 0.01) and intra-arterial adenosine (P < 0.01). Pretreatment with N(G)-nitro-l-arginine methyl ester or N-acetylpenicillamine did not affect the level of protection induced by rIPC (P = not significant, compared with rIPC alone) or intra-arterial adenosine (P = not significant, compared with intra-arterial adenosine alone), but prior femoral nerve transection or pretreatment with S-nitroso-N-acetylpenicillamine abolished the cardioprotective effect of intra-arterial adenosine and rIPC. Intra-arterial adenosine, like rIPC, releases a blood-borne cardioprotective factor(s) that is dependent on an intact femoral nerve and is inhibited by pretreatment with a NO donor. These results may be important when designing or assessing the results of clinical trials of adenosine or rIPC cardioprotection, where NO donors are used as part of therapy.