There is mixed evidence adenosine receptors (ARs) may enhance myocardial contractility, although this remains contentious. We assessed inotropic actions of adenosine (50 muM) and selective AR activation with 100 nM N (6)-cyclohexyladenosine (CHA; A(1)AR agonist), 25 nM 2-[p-(2-carboxyethyl) phenethylamino]-5'-N-ethylcarboxamidoadenosine (CGS-21680; A(2A)AR agonist) and 100 nM 2-chloro-N (6)-(3-iodobenzyl)-adenosine-5'-N-methyluronamide (Cl-IB-MECA; A(3)AR agonist) in mouse hearts perfused at constant pressure, constant flow, or conditions of stable flow and pressure (following maximal nitroprusside-mediated dilatation at constant flow). Adenosine and CGS-21680 significantly (although modestly) increased force in constant-pressure perfused hearts (=10 mmHg elevations in systolic pressure), effects paralleled by coronary vasodilatation (=10 ml min(-1) g(-1) elevations in flow). Neither CHA nor Cl-IB-MECA altered force or flow. With constant-flow perfusion, adenosine and CGS-21680 reduced systolic pressure in parallel with perfusion pressure. When changes in coronary flow and pressure were prevented, CGS-21680 failed to alter contractility. However, adenosine still enhanced systolic pressure up to 10 mmHg. Relations between flow, perfusion pressure and ventricular force evidence substantial Gregg effects in murine myocardium: systolic force increases transiently by approximately 1 mmHg ml(-1) min(-1) g(-1) rise in flow during the first minutes of hyperaemia and in a sustained manner (by approximately 1 mmHg mmHg(-1)) during altered perfusion pressure. These effects contribute to inotropism with AR agonism when flow/pressure is uncontrolled. In summary, we find no evidence of direct A(1) or A(3)AR-mediated inotropic responses in intact myocardium. Inotropic actions of A(2A)AR agonism appear entirely Gregg-related. Nonetheless, the endogenous agonist adenosine exerts a modest inotropic action independently of flow and perfusion pressure. The basis of this response remains to be identified.