The use of a reagent strip and reflectance meter for the bedside measurement of 3-hydroxybutyrate during the clinical management of diabetic ketoacidosis is described. Comparison of this method with a laboratory enzymatic assay shows good correlation (r = 0.97, p less than 0.05). Initial use in acute ketoacidosis suggests that knowledge of hourly changes in 3-hydroxybutyrate levels could be helpful in determining the optimum insulin dose.