1. Pregnancy in rats is characterized by a reduction in arterial pressure that is associated with a decreased response to vasoconstrictors. However, the responses to vasodilators in isolated vessels remain controversial and are not well established in hypertensive pregnant rats. 2. In the present study, we investigated the effect of pregnancy on the bradykinin (BK)-induced vasodilator responses of the isolated mesenteric arterial bed (MAB) from Wistar normotensive and spontaneously hypertensive rats (SHR) and determined the role of nitric oxide (NO), endothelium-derived hyperpolarizing factor (EDHF) and angiotensin-converting enzyme (ACE) in these responses. 3. Mean arterial pressure (MAP) in pregnant normotensive and pregnant hypertensive rats (93 +/- 1 and 122 +/- 2 mmHg, respectively) was lower than in non-pregnant controls (128 +/- 1 and 163 +/- 2 mmHg, respectively; P < 0.05). In MAB isolated from normotensive rats and precontracted with phenylephrine, the effects of bradykinin, acetylcholine (ACh) and nitroglycerine (NG) were not influenced by pregnancy. In contrast, the vasodilator responses to BK were significantly reduced in pregnant compared with non-pregnant SHR and seemed to be specific to BK. 4. The ACE inhibitor captopril potentiated BK vasodilator responses and abolished the differences between pregnant and non-pregnant SHR. Inhibition of nitric oxide (NO) synthase by N(G)-nitro-L-arginine methyl ester (l-NAME) significantly reduced the vasodilator effect of BK in all groups. In the presence of l-NAME plus high K+ solution (47 mmol/L), BK-induced vasodilation was completely blocked. The NO-dependent component of the responses seems to be more important in hypertensive rats and pregnancy does not modify this profile. 5. Our results suggest that increased ACE activity may be involved in the pregnancy associated reduction in vasodilator responses to BK in the MAB of hypertensive rats. Pregnancy does not modify the relative contribution of the EDHF and NO to the vasodilator effect of BK.