Intramural nerve stimulation elicited a powerful relaxation of the longitudinal muscle of the rabbit portal vein in the presence of atropine and guanethidine, but not of the guinea‐pig portal vein. Intramural nerve stimulation of the rabbit portal vein produced a 13 fold increase in release of 3H‐adenyl compounds after preloading with [3H]‐adenosine. About 50% of this release was abolished by guanethidine. All release was abolished by tetrodotoxin. No significant release of radioactive compounds was observed during intramural nerve stimulation of the guinea‐pig portal vein in the presence of guanethidine, although there was a 6 fold increase in release of radioactivity in the absence of drugs. Histochemical studies using quinacrine, which binds ATP showed a fine fluorescent nerve plexus, nerve bundles, and ganglion cells in the rabbit portal vein, but not in the guinea‐pig portal vein. This plexus was still present after chemical sympathectomy with 6‐hydroxydopamine. Adenosine 5′‐triphosphate (ATP) relaxed the rabbit portal vein, but usually produced a biphasic response, consisting of a contraction followed by a relaxation, of the guinea‐pig portal vein. Prostaglandins E1 and E2 caused contraction of the rabbit portal vein. Indomethacin, a prostaglandin synthesis inhibitor, potentiated the relaxations of the rabbit portal vein produced by both non‐adrenergic, non‐cholinergic nerve stimulation and ATP. High concentrations of antazoline and phentolamine, which antagonize purinergic responses in the guinea‐pig taenia coli, caused a loss of basal tone so that it was not possible to assess their effects on the responses of the portal vein to either non‐adrenergic, non‐cholinergic nerve stimulation, or ATP. Comparison of the results on the portal vein of the rabbit and guinea‐pig provides support for the view that: (i) quinacrine fluorescence can be used to localize purinergic nerves and that the rabbit portal vein is supplied by these nerves; (ii) ATP is released from adrenergic nerve fibres, although, based on histochemical analysis, about 3 to 7 times less than is released from purinergic nerve fibres. 1979 British Pharmacological Society