1. The function of inhibitory neuronal M2 muscarinic receptors in diabetic rat lungs was investigated. 2. Neuronal M2 muscarinic receptors inhibit acetylcholine release from parasympathetic nerves. Thus, stimulation of neuronal M2 muscarinic receptors with muscarinic agonists, such as pilocarpine, inhibits acetylcholine release and vagally induced bronchoconstriction. In contrast, blockade of neuronal M2 muscarinic receptors with selective M2 muscarinic antagonists, such as AF-DX 116, potentiates acetylcholine release and vagally induced bronchoconstriction. 3. Rats were made diabetic by streptozotocin (65 mg kg (-1), i.v.). After 7 14 days the rats were anaesthetized with urethane (1.5 g kg (-1), i.p.), tracheostomized, vagotomized, ventilated and paralysed with suxamethonium (30 mg kg (-1), i.v.). Some 7 day diabetic rats were treated with low doses of long acting (NPH) insulin (2 units day (-1), s.c.) for 7 days before experimentation. This dose of insulin was not sufficient to restore normoglycaemia in diabetic rats. Thus, insulin-treated diabetic rats remained hyperglycaemic. 4. Distal electrical stimulation (5 70 Hz, 6 s, 40 V, 0.4 ms) of the vagi caused bronchoconstriction, measured as an increase in inflation pressure and bradycardia. In diabetic rats, vagally induced bronchoconstriction was significantly depressed vs controls. In contrast, bronchoconstriction caused by i.v. acetylcholine was similar in diabetic and control animals. 5. The function of neuronal M2 muscarinic receptors was tested with the muscarinic agonist pilocarpine (0.001-100.0 microg kg (-1), i.v.) and the antagonist AF-DX 116 (0.01-3.0 mg kg (-1), i.v.). Pilocarpine inhibited vagally-induced bronchoconstriction (30 Hz, 20-40 V, 0.4 ms at 6 s) and AF-DX 116 potentiated vagally-induced bronchoconstriction (20 Hz, 20-40 V, 0.4 ms at 6 s) to a significantly greater degree in diabetic rats compared to controls. 6. Both frequency-dependent vagally-induced bronchoconstriction and M2 muscarinic receptor function could be restored to nearly control values in diabetic rats treated with low doses of insulin. 7. Displacement of [3H]QNB (1 nM) with the agonist carbachol (10.0 nM-10.0 mM) from diabetic cardiac M2 muscarinic receptors revealed a half log increase in agonist binding affinity at both the high and low affinity binding sites vs controls. In contrast, M2 receptors from insulin-treated diabetic rat hearts showed no significant difference in binding affinity vs controls. 8. These data show that neuronal M2 muscarinic receptors in the lungs have increased function in diabetic rats, suggesting that insulin modulates M2 muscarinic receptor function.