Approximately, a few percent of the European population suffers from diabetes. Scientific evidence showed that specific treatment of this disease could be successfully tailored on the basis of proper differential diagnosis that in many instances also requires genetic testing. This may be helpful in achieving metabolic control of the disease, increasing quality of life and potentially reducing the prevalence of chronic complications. Identification of the molecular background of these specific forms of diabetes gives new insight into the underlying aetiology. This knowledge helps to optimize treatment in specific clinical situations. Monogenic diabetes is an excellent example of a clinical area where new advances in molecular genetics can aid patient care and treatment decisions. The most frequently diagnosed forms of monogenic diabetes are MODY, mitochondrial diabetes, permanent and transient neonatal diabetes (PNDM and TNDM). These rare forms probably constitute at least a few percent of all diabetes cases seen in diabetic clinics. The proper differential diagnosis also helps to predict the progress of diabetes in affected individuals and defines the prognosis in the family. Recently, several genome wide association studies added new facts to the knowledge on complex forms of type 2 diabetes mellitus (T2DM) as the scientists substantially extended the short list of previously identified genes. Most newly identified variants influence beta-cell insulin secretion, while a few modulate peripheral insulin action. It is not clear whether in the future the genetic testing of frequent polymorphisms will influence the treatment of T2DM. In this review, we present the clinical application of genetic testing in non-autoimmune diabetes, mostly monogenic forms of disease.