This study analyses the relaxation induced by the hydroalcoholic extract of stems, leaves and roots from Phyllanthus urinaria (Euphorbiaceae) in the guinea-pig trachea (GPT) pre-contracted by carbachol. The hydroalcoholic extract of P. urinaria (0.1-10 mg mL-1) caused a graded relaxation in GPT with or without epithelium, with mean EC50 values of 1.94 (1.41-2.67) and 2.00 (1.47-2.78) mg mL-1 and Emax of 717 mg (+/- 16) and 627 mg (+/- 12), respectively. The relaxation in response to hydroalcoholic extract, like that to cromakalim (EC50 3.57 (2.75-4.64 microM) in GPT without epithelium, was fully abolished in the presence of high KCl concentrations (80 mM), and was significantly attenuated by tetraethylammonium (10 or 30 mM) or glibenclamide (0.1 or 3 microM). However, the relaxation caused by the hydroalcoholic extract was unaffected by apamin (0.1 or 1.0 microM), nitro-L-arginine (L-NOARG, 100 microM), methylene blue (10 microM) or by calcitonin gene-related peptide (CGRP) (8-37) (a CGRP antagonist, 0.1 microM). Both propranolol (1 or 3 microM) and [D-p-Cl-Phe6,Leu17]VIP (a vasoactive intestinal peptide (VIP) receptor antagonist, 0.1 microM) produced a significant displacement to the right (about 2-fold) of the relaxation response to hydroalcoholic extract of P. urinaria. Thus, the present results indicate that the ATP-activated potassium channels sensitive to glibenclamide, but not the small conductance calcium-activated potassium channels sensitive to apamin, largely contribute to the relaxation effect of the hydroalcoholic extract of P. urinaria in GPT. In addition, both beta 2 and VIP-mediated responses seem to account, at least in part, for the relaxation effect of the hydroalcoholic extract, as its relaxant response was partially attenuated by both propranolol and VIP receptor antagonist.