The concept of diabetic retinopathy as a microvascular disease has evolved and is now considered a more complex diabetic complication in which neurovascular unit impairment plays an essential role and, therefore, can be considered as a main therapeutic target in the early stages of the disease. However, neurodegeneration is not always the apparent primary event in the natural story of diabetic retinopathy, and a phenotyping characterization is recommendable to identify those patients in whom neuroprotective treatment might be of benefit. In recent years, a myriad of treatments based on neuroprotection have been tested in experimental models, but more interestingly, there are drugs with a dual activity (neuroprotective and vasculotropic). In this review, the recent evidence concerning the therapeutic approaches targeting neurovascular unit impairment will be presented, along with a critical review of the scientific gaps and problems which remain to be overcome before our knowledge can be transferred to clinical practice.