Bradykinin and histamine have been shown to act as vasodilators via endothelial nitric oxide synthesis and soluble guanylyl cyclase activation in the underlying vascular smooth muscle. As endothelial guanylyl cyclase can also be activated, this could be the mechanism for increased capillary permeability via cGMP formation and PKG activation. We have previously shown that histamine signals permeability increase in this way in pial venular capillaries, but bradykinin uses a different pathway that requires free radicals. We now present data that indicate the bradykinin pathway is not confined to the blood-brain barrier. The ileal artery of a freshly killed Wistar rat was cannulated orthogradely and branches that did not lead to the cremaster muscle were ligated. The cremaster microcirculation was flushed with St Thomas’ cardioplegic solution containing heparin (300 IU ml^-1). The scrotum was cut to reveal the cremaster and testes. An incision was made to open the cremaster, care being taken not to damage any major vessel. The muscle was spread over a transparent support and superfused with a Krebs buffer solution maintained at 37°C. The perfusate was changed to a buffer containing albumin (10 mg ml^-1) with added FITC-albumin (5 mg ml^-1). Perfusate flow was stopped and pressure differences in the vasculature were allowed to dissipate. Permeability was obtained from the rate of change in fluorescent signal across the wall of a selected venule (k) and its diameter (d): P = kd/4. The resting albumin permeability was 0.30 ± 0.05 x 10^-6 cm s^-1(mean ± sem, n = 36), which was increased to 2.4 ± 0.22 and 3.6 ± 1.0 (all n = 4) with 1 µM bradykinin and histamine, respectively. The response to bradykinin was blocked by scavenging free radicals with superoxide dismutase and catalase (100 U ml^-1 each; 0.11 ± 0.05), but unaffected by the nitric oxide synthase inhibitor L-NAME (100 µM; 1.9 ± 0.36). On the other hand the response to histamine was unaffected by superoxide dismutase and catalase (2.0 ± 0.25), but blocked by L-NAME (0.07 ± 0.03). This indicates that the signal transduction in this skeletal muscle is similar to that of the blood-brain barrier. L-NAME application alone increased permeability slightly (0.36 ± 0.07, p < 0.05 paired t test), but when co-applied with superoxide dismutase and catalase this permeability change was blocked (-0.10 ± 0.12), which indicates that constitutive NO reduces permeability by scavenging free radicals.