Pain and itch sensations are induced by depolarization of C-fibre nerves and possibly other types of fibres. We have evidence from several species, including mice, that skin plasma extravasation induced by the Phoneutria nigriventer spider venom (PNV) is dependent on tachykinin NK(1) receptors. We have now investigated the itching measured as bouts of scratching in response to intradermal (i.d.) PNV in wildtype (NK(1)(+/+)) and NK(1) receptor knockout (NK(1)(-/-)) mice. Mice, either NK(1)(+/+) or NK(1)(-/-), were given a single i.d. injection (0.05 ml) of test agent or vehicle into the shaved dorsal skin, in the intercostal region, in a randomized way. The bouts of scratching were recorded in a blinded manner for 60 min. Oedema formation was concomitantly assessed by the extravascular accumulation of i.v. injected (125)I-albumin. The i.d. injection of either substance P (at a high dose of 100 nmol/site), or PNV (0.3-10 microg/site) induced oedema formation in NK(1)(+/+) but substantially less was observed in NK(1)(-/-) mice, as previously reported. PNV also induced scratching, but significantly less scratching was observed in NK(1)(-/-) compared with NK(1)(+/+) mice. In contrast, SP did not induce significant scratching at amounts up to 100 nmol in NK(1)(+/+) mice. Experiments with an NK(1) receptor antagonist SR140333 (at doses that blocked PNV-induced oedema) revealed that whilst a local co-injection i.d. (1 nmol) in NK(1)(+/+) mice had no effect on PNV (3 microg/site)-induced scratching (18.5+/-3.7 vs. 14.4+/-3.5 bouts, mean+/-S.E.M., n=5-7), systemic treatment with SR140333 (120 nmol/kg, i.v.) significantly inhibited scratching (14+/-3.5 vs. 3.1+/-1.2 bouts, n=4-6; P<0.05). These results indicate that NK(1) receptors are involved in mediating PNV-induced scratching and that the location of the receptors is unlikely to be skin. Thus, a distinct separation between endogenous microvascular and PNV nociceptive NK(1)-dependent effects is suggested.