Based upon clinical observation of undue prolongation of pancuronium- induced blockade in the presence of intravenous infusion of nitroglycerin, neuromuscular blockades produced by pancuronium, succinylcholine and d-tubocurarine were studied in 51 cats using the sciatic–gastrocnemius nerve–muscle preparation. Pancuronium-induced blockade was found to be significantly prolonged (P < .01) in the presence of a nitroglycerin infusion of 1 µg/kg/min (65 vs. 127 min). Less, but still significant, prolongation occurred when nitroglycerin, 0.5 µg/kg/min, was infused. The intravenous infusion of nitroglycerin must be started prior to the pancuronium injection for the block to be prolonged. Neuromuscular blocks produced by succinylcholine and d-tubocurarine were not altered by nitroglycerin. In experiments using the isolated rat diaphragm preparation, the depth of pancuronium-induced block was found not to be changed by nitroglycerin, suggesting an effect of nitroglycerin on the process of recovery from blockade. These findings indicate a selective pancuronium-nitroglycerin interaction.