Patients with Type 2 diabetes mellitus (T2DM) frequently exhibit a distinctive cardiac phenotype known as diabetic cardiomyopathy. Cardiac complications associated with T2DM include cardiac inflammation, hypertrophy, fibrosis and diastolic dysfunction in the early stages of the disease, which can progress to systolic dysfunction and heart failure. Effective therapeutic options for diabetic cardiomyopathy are limited and often have conflicting results. The lack of effective treatments for diabetic cardiomyopathy is due in part, to our poor understanding of the disease development and progression, as well as a lack of robust and valid preclinical human models that can accurately recapitulate the pathophysiology of the human heart. In addition to cardiomyocytes, the heart contains a heterogeneous population of non-myocytes including fibroblasts, vascular cells, autonomic neurons and immune cells. These cardiac non-myocytes play important roles in cardiac homeostasis and disease, yet the effect of hyperglycaemia and hyperlipidaemia on these cell types are often overlooked in preclinical models of diabetic cardiomyopathy. The advent of human induced pluripotent stem cells provides a new paradigm in which to model diabetic cardiomyopathy as they can be differentiated into all cell types in the human heart. This review will discuss the roles of cardiac non-myocytes and their dynamic intercellular interactions in the pathogenesis of diabetic cardiomyopathy. We will also discuss the use of sodium-glucose cotransporter 2 inhibitors as a therapy for diabetic cardiomyopathy and their known impacts on non-myocytes. These developments will no doubt facilitate the discovery of novel treatment targets for preventing the onset and progression of diabetic cardiomyopathy.