The current classification of Diabetes mellitus is well established in the healthcare system but does not mirror the heterogeneity of the disease. Using cluster analysis, a Swedish group suggested a novel classification dividing Diabetes mellitus into five subgroups. These subgroups differ substantially in prediction and therapy of long-term diabetes complications. Among others HbA (1c) , GAD-autoantibodies, HOMA2-B and HOMA2-IR were used for the classification. The calculation of the HOMA indices relies on the values of glucose and insulin or C-peptide, respectively. Laboratory parameters need to be comparable i.e. the measurement procedures need to be internationally standardized to use this classification world-wide. Independent from this classification comparability of laboratory results is necessary for diagnosis, prediction and therapy of people with diabetes. In recent decades considerable efforts were made to reach international standardization and comparability. During the last decade, measurement procedures for glucose and HbA (1c) were standardized. The standardization of C-peptide measurements should be finished by 2020 and implemented thereafter. In contrast, the standardization of insulin measurements is less advanced. GAD-autoantibodies cannot be standardized because of their biological heterogeneity. However, harmonization has been achieved in the last years. The world-wide standardization of the aforementioned laboratory parameters can reduce parameter-based wrong diagnosis and wrong clinical decisions for the benefit of the patient.