1 The radioactive microsphere method was used to study the effects of ergotamine (5, 10 and 20 mug/kg, i.v.) on systemic and regional haemodynamic variables in chloralose-urethane anaesthetized cats. The influence of the drug was also studied on the number of 15 mum microspheres escaping entrapment in the head to emerge in the left external jugular vein.2 Ergotamine decreased the heart rate and cardiac output. Since arterial blood pressure remained unchanged, calculated total peripheral resistance increased.3 The regional distribution of cardiac output obtained with 15 mum microspheres agreed well with previous studies in cats where 25 mum spheres were used. The most pronounced difference was that in the present investigation more microspheres, apparently escaping through arteriovenous anastomoses (AVAs), were detected in the lungs than when larger spheres had been used.4 Coronary blood flow decreased, while uterine blood flow was increased by the drug. The microsphere content of the lungs, which receive the spheres not only via bronchial arteries but also via AVAs, was greatly reduced by all doses of ergotamine. Ergotamine did not influence tissue blood flow to other major organs such as the brain, kidneys, skin, liver, skeletal muscle or the gastrointestinal tract.5 In the 16 experiments, 0.46 +/- 0.05 (s.e. mean)% of the total microspheres injected (equivalent to 11.7 +/- 1.4% of microspheres detected in the left-side of the head) appeared within 2 min of microsphere injection into the left external jugular vein. The highest dose of ergotamine significantly reduced the shunting of the microspheres in the head.6 Since 15 mum microspheres are only likely to reach the lungs by passing into the venous circulation through large glomus-type AVAs, we conclude that ergotamine reduces the fraction of microspheres appearing in the lungs by causing strong vasoconstriction in the AVAs in the head.7 In conformity with the closure of head AVAs is the finding that ergotamine reduced the jugular venous Po(2) and O(2) saturation thereby increasing the A-V O(2) saturation difference.8 It is quite possible that decreased A-V shunting may be the prominent mechanism of the antimigraine action of the drug, since sudden opening of AVA's has been implicated in the pathophysiology of migraine-syndrome.