Empagliflozin (EMPA), a potent inhibitor of the renal sodium-glucose cotransporter 2 and an effective treatment for Type 2 diabetes, has been shown to have cardioprotective effects, independent of improved glycaemic control. Several non-canonical mechanisms have been proposed to explain these cardiac effects, including increasing circulating ketone supply to the heart. This study aims to test whether EMPA directly alters cardiac ketone metabolism independent of supply. The direct effects of EMPA on cardiac function and metabolomics were investigated in Langendorff rat heart perfused with buffer containing 5 mM glucose, 4 mM β-hydroxybutyrate (βHb) and 0.4 mM intralipid, subject to low flow ischaemia/reperfusion. Cardiac energetics were monitored in situ using 31P NMR spectroscopy. Steady-state 13C labelling was performed by switching 12C substrates for 13C1 glucose or 13C4 βHb and 13C incorporation into metabolites determined using 2D 1H-13C HSQC NMR spectroscopy. EMPA treatment improved left ventricular-developed pressure during ischaemia and reperfusion compared to vehicle-treated hearts. In EMPA-treated hearts, total adenosine triphosphate (ATP) and phosphocreatine (PCr) levels, and Gibbs free energy for ATP hydrolysis were significantly higher during ischaemia and reperfusion. EMPA treatment did not alter the incorporation of 13C from glucose into glycolytic products lactate or alanine neither during ischaemia nor reperfusion. In ischaemia, EMPA led to a decrease in 13C1 glucose incorporation and a concurrent increase in 13C4 βHb incorporation into tricarboxylic acid (TCA) cycle intermediates succinate, citrate, and glutamate. During reperfusion, the concentration of metabolites originating from 13C1 glucose was similar to vehicle but those originating from 13C4 βHb remained elevated in EMPA-treated hearts. Our findings indicate that EMPA causes a switch in metabolism away from glucose oxidation towards increased ketone utilization in the rat heart, thereby improving function and energetics both during ischaemia and recovery during reperfusion. This preference of ketone utilization over glucose was observed under conditions of constant supply of substrate, suggesting that EMPA acts directly by modulating cardiac substrate preference, independent of substrate availability. The mechanisms underlying our findings are currently unknown, warranting further study.