In both Type 1 and Type 2 diabetes, there is a gradual loss of pacreatic beta cells with time. The degree of beta cell destruction may be indirectly measured by the appearance of islet-specific components in circulation. For instance, insulin mRNA is increased in circulation in T1D patients before clinical loss of function of transplanted islets. We have recently demonstrated, that hypomethylated insulin gene characteristic for beta cells leaks into circulation during beta cell death and can predict hyperglycemia in both STZ-induced and spontaneous (NOD) mouse models. In pre-diabetic NOD mice, hypomethylated proinsulin DNA in circulation negatively correlated with total pancreatic insulin content measured in acid-ethanol extracts. However, in a natural history of prediabetes, the destruction may be slow at the very beginning due to low numbers of effector T cells, or be slow immediately before clinical precipitation of diabetes, because most of the target beta cell mass has been lost already. Thus the same level of destruction may have different prognostic value. The only accurate way to interpret the level of destruction would be to know residual beta cell mass. For that, in clinical setting, surrogate markets have been traditionally used, such as C-peptide, or, more recently, beta cell mass have been visualized by positron emission tomography. However, in humans, this data has never been directly correlated to total pancreatic insulin content. We hypothesize that circulating C-peptide (amylin, proinsulin) levels strongly correlate with total pancreatic insulin content, and that a ratio between the biomarker of beta cell death and C-peptide determines prognosis.