The purpose of this study was to explore accuracy in use of continuous glucose monitoring (CGM) together with oral glucose tolerance test (OGTT) to detect impaired glucose tolerance and early onset of type 1 diabetes and to define its stages during a follow-up of children with increased genetic susceptibility to type 1 diabetes. Follow-up studies have reported that children with multiple islet autoantibodies have a lifetime risk of 85% to develop type 1 diabetes and its prediction has turned out to be challenging and new methods to prediction are needed. Study subjects were recruited from the Finnish type 1 diabetes prediction and prevention (DIPP) study. Subjects were divided into five groups depending on their islet autoantibody profile and glucose tolerance defined on baseline. Each subject went trough 7-10 days CGM period which included OGTT. If wanted, subjects performed OGTT also at home and CGM curves from Home-OGTT were compared to see if OGTT would be accurate to perform remotely if needed. The CGM curves during OGTT were analyzed and compared to plasma glucose values taken during OGTT. Results show that CGM can reliably detect asymptomatic stage 3 diabetes during a standard OGTT and detect a significant difference during a 10-day sensoring period compared to non-diabetic individuals. These results show that CGM could be an alternative tool to predict early onset of type 1 diabetes. When compared the CGM curve to plasma glucose curve during OGTT, the peak appeared approximately 15 min later with CGM than with plasma glucose, when the most common time of the peak was at 30 minutes timepoint. This can be due to the lag time with CGM, but it would be interesting to see the 45 min timepoint with plasma glucose to see if the peak seen with CGM is more reliable. Generally, CGM provides more information about subjects glucose metabolism compared to OGTT, confirms normoglycemia and detects dysglycemia and asymptomatic stage 3 diabetes.