Vasodilator-induced elevation of intracellular cyclic AMP (cAMP) is a central mechanism governing arterial relaxation but is incompletely understood due to the diversity of cAMP effectors. Here we investigate the role of the novel cAMP effector exchange protein directly activated by cAMP (Epac) in mediating vasorelaxation in rat mesenteric arteries. In myography experiments, the Epac-selective cAMP analogue 8-pCPT-2-O-Me-cAMP-AM (5 μM, subsequently referred to as 8-pCPT-AM) elicited a 77.6 ± 7.1% relaxation of phenylephrine-contracted arteries over a 5 min period (mean ± SEM; n = 6). 8-pCPT-AM induced only a 16.7 ± 2.4% relaxation in arteries pre-contracted with high extracellular K(+) over the same time period (n = 10), suggesting that some of Epac's relaxant effect relies upon vascular cell hyperpolarization. This involves Ca(2+)-sensitive, large-conductance K(+) (BK(Ca)) channel opening as iberiotoxin (100 nM) significantly reduced the ability of 8-pCPT-AM to reverse phenylephrine-induced contraction (arteries relaxed by only 35.0 ± 8.5% over a 5 min exposure to 8-pCPT-AM, n = 5; P < 0.05). 8-pCPT-AM increased Ca(2+) spark frequency in Fluo-4-AM-loaded mesenteric myocytes from 0.045 ± 0.008 to 0.103 ± 0.022 sparks s(-1) μm(-1) (P < 0.05) and reversibly increased both the frequency (0.94 ± 0.25 to 2.30 ± 0.72 s(-1)) and amplitude (23.9 ± 3.3 to 35.8 ± 7.7 pA) of spontaneous transient outward currents (STOCs) recorded in isolated mesenteric myocytes (n = 7; P < 0.05). 8-pCPT-AM-activated STOCs were sensitive to iberiotoxin (100 nM) and to ryanodine (30 μM). Current clamp recordings of isolated myocytes showed a 7.9 ± 1.0 mV (n = 10) hyperpolarization in response to 8-pCPT-AM that was sensitive to iberiotoxin (n = 5). Endothelial disruption suppressed 8-pCPT-AM-mediated relaxation in phenylephrine-contracted arteries (24.8 ± 4.9% relaxation after 5 min of exposure, n = 5; P < 0.05), as did apamin and TRAM-34, blockers of Ca(2+)-sensitive, small- and intermediate-conductance K(+) (SK(Ca) and IK(Ca)) channels, respectively, and N(G)-nitro-L-arginine methyl ester, an inhibitor of nitric oxide synthase (NOS). In Fluo-4-AM-loaded mesenteric endothelial cells, 8-pCPT-AM induced a sustained increase in global Ca(2+). Our data suggest that Epac hyperpolarizes smooth muscle by (1) increasing localized Ca(2+) release from ryanodine receptors (Ca(2+) sparks) to activate BK(Ca) channels, and (2) endothelial-dependent mechanisms involving the activation of SK(Ca)/IK(Ca) channels and NOS. Epac-mediated smooth muscle hyperpolarization will limit Ca(2+) entry via voltage-sensitive Ca(2+) channels and represents a novel mechanism of arterial relaxation.