The energetic requirements of the heart are weight-for-weight higher than for any other organ, providing non-stop function for a lifetime while maintaining reserve in response to increased demand. This is met by continuously recycling a relatively small pool of adenosine triphosphate (ATP), which when hydrolysed to ADP liberates ‘free energy’ available to perform work (ΔGATP). This process is aided by mitochondrial-creatine kinase (CK) located at the mitochondria which catalyses the transfer of a high-energy phosphoryl from ATP onto creatine to form ADP and phosphocreatine (PCr), and by CK coupled to ATPases which catalyses the reverse reaction. This system functions to maintain local ATP/ADP ratios, provide rapid transfer of chemical energy, and as an energy buffer, with PCr available for rapid regeneration of ATP when demand outstrips supply. In the failing heart, total creatine levels, PCr, and CK activity are all down-regulated, regardless of species or aetiology, and eventually even ATP levels are reduced. This has led to the concept of the failing heart as ‘energy starved’, but whether these energetic changes are biomarkers or significantly contribute to the pathophysiology of heart failure is the subject of continued debate. Key methodologies are now becoming available in vivo to help address this question in rodent models. For example, 31P magnetic resonance spectroscopy (MRS) can be used to detect ATP, PCr, and measure CK flux using saturation transfer protocols, while 1H-MRS can detect total creatine. In addition, recent advances have drastically shortened scan times for anatomical MRI in rodents, enabling multi-parametric MR under the same general anaesthesia. A major challenge will be the combination of these approaches to calculate ΔGATP in vivo, which represents the ultimate integrated measure of energetic status. These techniques have been applied to genetic mouse models of impaired creatine biosynthesis and CK ablation, which have served to highlight remarkable plasticity in cardiac energy metabolism, and confirm the importance of the CK system under conditions of high workload and acute stress. Surgical models of rodent heart failure develop the same hallmark defects in cardiac energetics and have proven to be particularly informative. Key factors affecting interpretation of these models will be discussed, as will recent studies demonstrating beneficial effects from therapeutic targeting of the CK system. It is anticipated that these important new findings will stimulate interest in metabolic therapies and drive technological advances aimed at investigating cardiac energetics in vivo.