1. We have assessed the effect of allopurinol, amlodipine and propranolol pretreatment on both endothelium-dependent and endothelium-independent coronary vasodilatation in vivo, by comparing pre-ischaemic responses with those measured after 60 min of coronary artery occlusion and 30 min of reperfusion in anaesthetized dogs. 2. In 15 untreated dogs ischaemia and reperfusion attenuated the increases in coronary blood flow produced by either acetylcholine (0.01-0.05 micrograms kg-1, i.a.) or glyceryl trinitrate (0.05-0.2 micrograms kg-1, i.a.), to an average of 39 +/- 4% and 42 +/- 5% of the pre-ischaemic control response, respectively (both P < 0.05). 3. In 5 dogs treated with allopurinol (25 mg kg-1, orally, 24 h previously, plus 50 mg kg-1, i.v., 5 min before occlusion), the increases in coronary blood flow after ischaemia and reperfusion (acetylcholine: 78 +/- 12%, glyceryl trinitrate: 60 +/- 3% of pre-ischaemic response) were significantly larger than post-ischaemic responses in untreated dogs (both P < 0.05). 4. Similarly, amlodipine treatment (3 micrograms kg-1 min-1, i.v., starting 90 min before occlusion) in 5 dogs improved post-ischaemic increases in blood flow (acetylcholine: 58.5%, glyceryl trinitrate: 66 +/- 6% of pre-ischaemic response, significantly greater than post-ischaemic responses in untreated dogs, P < 0.05). 5. In contrast, in a further 6 dogs pretreated with propranolol (1 mg kg-1, i.v., 30 min before occlusion,plus 0.5 mg kg-1 h-1, i.v.), blood flow responses after ischaemia and reperfusion were not different from post-ischaemic responses in untreated dogs (acetylcholine: 46 +/- 6%, glyceryl trinitrate: 46 +/-6% of pre-ischaemic response).6. These results suggest that allopurinol and amlodipine protect against the post-ischaemic impairment of endothelium-dependent and endothelium-independent coronary vasodilatation in vivo by mechanisms additional to endothelial protection.