Pulsatile insulin release from glucose-stimulated b-cells is controlled by synchronized oscillations of the sub-plasma membrane concentrations of Ca²⁺ and cAMP, but little is known about the mediators of cAMP action. The aim of this study was to clarify the contribution of protein kinase A (PKA) and Exchange protein activated by cAMP (Epac) to pulsatile insulin secretion. Fluorescent dyes and translocation biosensors were used in combination with evanescent wave microscopy to record Ca²⁺ or cAMP in parallel with the membrane phospholipid PIP3, which is formed by autocrine insulin receptor activation and is a useful readout for insulin secretion from single MIN6 b-cells. Elevation of the glucose concentration from 3 to 11 mM induced oscillations of Ca²⁺ and cAMP that were tightly associated with PIP3 oscillations, reflecting pulsatile release of insulin. In cells exposed to the PKA inhibitors Rp-8-CPT-cAMPS or KT5720, the PIP3 response to subsequent glucose stimulation was reduced by 30±7% (n=37, P<0.01) and 59±10% (n=19, P<0.01), respectively, despite unaltered Ca²⁺ signals. In contrast, already manifested PIP3 oscillations were unaffected by PKA inhibition. Nevertheless, cAMP was required also for established pulsatile insulin release, since inhibition of adenylyl cyclases with 2',5'-dideoxyadenosine or SQ22,536 reduced the PIP3 response by 66±5% (n=54, P<0.01) and 42±3% (n=101, P<0.01), respectively. This inhibition was reversed and PIP3 oscillations restored by 1 mM of the Epac-selective nucleotide analogue 8-pMeOPT-cAMPS-AM. It is concluded that appropriate initiation of insulin release in glucose-stimulated b-cells requires PKA, while the cAMP-dependence of maintained pulsatile secretion is mediated by Epac.