1. The effects of the non-selective endothelin (ET) receptor (ETA/ETB) antagonist, bosentan, on sciatic nerve dysfunction in experimental diabetes were investigated. 2. Rats with 5-6 weeks untreated streptozotocin-diabetes exhibited characteristic slowed motor nerve conduction velocity (mean +/- s.d., 36.6 +/- 3.4 m s-1) and nerve laser Doppler flux (197 +/- 64 arbitrary units) compared to age-matched control animals (42.7 +/- 2.4 m s-1 and 398 +/- 77 arbitrary units, respectively). Preventative treatment of diabetic rats with bosentan at 100 mg kg-1 day-1 p.o. attenuated both these deficits (39.7 +/- 3.0 m s-1 and 305 +/- 56 arbitrary units, respectively) without affecting mean arterial pressure. 3. In control and untreated diabetic rats, ET-1, 1 nmol kg-1 i.v., caused an initial hypotension (duration, 30 +/- 13 and 26 +/- 9 s, respectively; change in mean arterial pressure, -27 +/- 13 and -25 +/- 7 mmHg, respectively) followed by prolonged hypertension (change in mean arterial pressure, 52 +/- 18 and 31 +/- 5 mmHg, respectively). Effectiveness of the chronic bosentan treatment was demonstrated by inhibition of the hypotensive response to ET-1 in treated diabetic rats (duration, 5 +/- 2 s; change in mean arterial pressure, -4 +/- 2 mmHg) although the hypertension was unaltered (change in mean arterial pressure, 32 +/- 9 mmHg). 4. Acute i.v. administration of 10 mg kg-1 bosentan caused variable and transient rises in nerve laser Doppler flux in control (78 +/- 63 arbitrary units) and untreated diabetic rats (93 +/- 77 arbitrary units). Acute bosentan blocked the hypotensive response to subsequent ET-1 administration and attenuated the later hypertension (change in mean arterial pressure, 21 +/-9 mmHg in control, 29 +/- 10 mmHg in diabetic).5. Our results indicate that oral treatment of diabetic rats with an ET receptor antagonist can improves ciatic nerve perfusion and conduction, suggesting that the vasoconstrictor action of endogenous ET may contribute to peripheral nerve dysfunction in experimental diabetes.