A simplified model of the blood-glucose regulatory system has been conformed to data from glucose-tolerance tests in nondiabetics measured by continuous sampling after oral ingestion and by intermittent determinations after different routes of administration. Despite the numerous approximations inherent in the model, all the data were conformed within the limits of experimental error. This extended the applicability of the simplified model used and indicated that the assumption of a single regulatory hormone such as insulin was adequate to fit most normal responses. Small deviations between some of the simulated curves and the actual data suggested modifications that would improve the model. Such changes would sacrifice simplicity for greater accuracy in predicting the detailed response.