Cyclic AMP production within the cardiac cell is highly compartmentalised allowing different receptors which increase cAMP to produce distinct cellular responses. Our recent work has shown that the organisation of signal components within microdomains of the cell membrane makes a significant contribution to spatial cAMP control [1,2]. Lipid rafts are liquid-ordered domains of the membrane enriched in cholesterol and sphingolipids. Caveolae (‘little caves’) are invaginated rafts defined by the presence of caveolin (Cav) and cavin proteins. Clustering of elements of particular signal cascades within a caveola promotes efficiency and fidelity of signalling, and caveolin provides additional control through its regulatory interaction with many protein partners via a 20 residue caveolin scaffolding domain (CSD). Although caveolae and non-caveolar rafts co-exist, evidence suggests that most proteins are clustered by caveolae in the cardiac cell [3]. In the adult rat ventricular myocyte, stimulation of β1- and β2-adrenoceptors (AR) and the type E prostaglandin receptor (EPR) increases total cellular cAMP (indexed using a cytosolic FRET cAMP biosensor based on Epac2). However activation of these receptors produces diverse functional responses. β1-AR stimulation increases ICa,L and [Ca2+]i and has pronounced positive inotropic and lusitropic effects. By contrast, β2-AR stimulation has a minimal impact on these parameters and EPR stimulation is without effect. Clearly, the cAMP signal produced by stimulation of these 3 receptors has access to different subcellular sites. The inotropic and lusitropic response to β1-AR stimulation can be directly linked with a large and sustained increase in cAMP seen specifically in regions where protein kinase A (PKA) type RII is tethered to A kinase anchoring proteins (assessed using a PKA-based sensor). Only transient changes in cAMP were recorded in this compartment following β2-AR and EPR stimulation. Disruption of caveolae with the cholesterol-depleting agent methyl-β-cyclodextrin (MBCD) enhanced PKA probe responses to β1- and β2-AR (but not EPR) stimulation but was without effect on Epac2 probe responses. Again, the change in PKA probe response with MBCD was directly correlated with functional changes: enhanced inotropic and lusitropic responses to stimulation of both β-AR subtypes were seen. As β2-AR are predominantly caveolar, β1-AR are found in both caveolar and non-caveolar fractions, whereas EPR are excluded from caveolar fractions, these data suggest that disruption of caveolae selectively promotes production of cAMP in a PKA RII domain through effects on receptors located in caveolae (i.e. all β2-AR and a subset of β1-AR). A simple explanation for this is that MBCD removes the normal inhibitory effect of Cav3 on adenylyl cyclase (AC) thereby promoting cAMP production. Interestingly, the cAMP signal revealed by caveolar disruption has access to different sites depending on the β-AR receptor activated. With β1-AR stimulation MBCD treatment increases ICa,L and hastens relaxation and [Ca2+]i transient decay, consistent with targeting of both the Ca2+ channel and phospholamban (PLB) by the enhanced cAMP signal. By contrast, the nascent β2-AR cAMP signal seen following MBCD treatment only targets PLB; no increase in ICa,L or phosphorylation of TnI or RyR are observed. The marked increase in phosphorylation of PLB (pPLB) evoked by β2 AR stimulation in MBCD-treated cells is mimicked by selective disruption of Cav3 binding using a cell-permeable Cav3 CSD peptide, consistent with a caveolae-specific effect of MBCD. Although AC activity may increase if the inhibitory influence of Cav3 is removed, our data suggest that the main factor contributing to the caveolar compartmentation of β2-AR signalling is facilitation of coupling of the β2-AR with Gi which inhibits cAMP production [4]. However, this does not explain the selective control of pPLB. We believe that this can be accounted for by the fact that phosphorylation of PLB is selectively restrained by PP1 tethered along with its inhibitor I-1 to AKAP18δ [5]. The sustained (and propagating) cAMP signal seen when β2-AR-Gi coupling is severed by caveolar disruption acts as a switch to turn off PP1 by PKA-phosphorylation of I-1, thereby allowing phosphorylation of PLB to be maintained. In summary, our work shows that compartmentation of cAMP signalling can be explained, in part, by the organisation of proteins in membrane microdomains. Cavaolae normally act to restrain the cAMP signal specifically in PKA RII domains following stimulation of caveolar β1- and β2-ARs. Given this, we predict that alterations in caveolae seen in disease (e.g. heart failure) and with treatment for disease (e.g. statins) will make a significant contribution to spatial control of cAMP and cardiac function during sympathetic stimulation.