Secretory diarrhoea is thought to arise from secretion by intestinal enterocytes (Field et al., 1978) but evidence from in vivo studies is ambiguous (Lucas, 2010). We have investigated the alternative mechanism of increased mucosal blood pressure causing fluid exudation into the lumen. Anaesthesia was induced in 22 male rats (250-400g), fasted overnight, with 80mg /kg pentobarbitone sodium I.P. and maintained with 24mg/ kg I.P. as required. The surgical procedure consisted of tracheostomy, carotid artery cannulation to record blood pressure (ABP), jugular/iliac vein cannulations for infusions, bilateral vagotomy to abolish vago-vagal reflexes, ligation of the duodenum to prevent entry of chyme and cannulation of a 30cm jejunal loop. Body temperature was maintained at 37°C. After experiment, euthanasia was by 100mg Euthatal I.V. Jejunal loops were recirculated with 40.0 ml 154mM choline chloride solution containing 100μM 5(N-ethyl, N-isopropyl) amiloride to prevent Na+ uptake, at 37°C and pH 7.4 at a rate of 1.0ml/min for at least 1h after which perfusate was recovered and weighed to assess secretion. This was determined in experiments in which ABP was shifted to a new steady level during the perfusion period by I.V. infusions of vasoactive agents, each acting through a different receptor system. The control experiments (Fig. 1) showed secretion of +2.2 ± 5.3 μl/cm/h . However, this masked a significant inverse relationship between secretion and diastolic ABP (r sq. = 72%, P slope < 0.001). Slow I.V. infusion of 1-2 I.U. arginine vasopressin (AVP) elevated DABP and caused absorption. Slow I.V. infusion of 10-50 μg VIP or 100-300 μg acetyl-β-methylcholine (MC) or bolus injection of 200-500 μg phentolamine to reduce the DABP to 80-40mmHg caused secretion as high as +50 μl/cm/h. When the agents further reduced DABP to below 40mmHg, net absorption was consistently found: secretion was then re-instituted by restoring DABP in the case of phentolamine through infusing AVP. Light histological examination showed normal mucosal morphology. Secretion versus DABP (Fig.1) or mean ABP were well fitted by a cubic equation (r sq = 72%, P slope < 0.001). While it is possible that VIP and MC had a secretagogue component in addition to their vasodepressor action, this could not have been the case with phentolamine. Accordingly, intestinal fluid flux is strongly dependent on DABP since, in the range of 40-80mmHg, resultant net secretion of appreciable magnitude occurs. In the context of the Starling forces across the capillary bed, our results could be explained if reduced peripheral resistance caused by the vasodilators led to increased perfusion of the mucosal capillary beds. This would be consistent with increased capillary hydrostatic pressure causing net outward filtration of fluid leading to net secretion into the small intestinal lumen.