The plasma glucose and insulin response to a standardized meal test breakfast was compared with the time-honored glucose tolerance test in the same normal healthy subjects. The amplitude of glycemic excursion and between-subject variation was less with the more physiologic standardized test meal than with that seen with the glucose tolerance test. The glucose tolerance test's prime function is to amplify any glucose intolerance, thus aiding diagnosis, whereas a standardized meal gives a more clinically relevant metabolic status. The administration of serial test meals during the same day in a smaller group of normal subjects indicated, as seen previously with repeated glucose tolerance tests, a diminishing carbohydrate tolerance during the day.