The objective of this study was to determine the relationship between beta-hydroxybutyrate levels and seizure control in children on the ketogenic diet. Seventy-four children on the ketogenic diet presenting for routine follow-up visits had blood levels of beta-hydroxybutyrate correlated with their seizure control. Forty-two children admitted for initiation of the ketogenic diet had urine ketones measured by dipstick and correlated with simultaneous blood levels of beta-hydroxybutyrate. Blood beta-hydroxybutyrate levels statistically correlated with seizure control (P = .003). Children with blood beta-hydroxybutyrate levels greater than 4 mmol/L were significantly more likely to have a decrease in seizure frequency than those with levels less than 4 mmol/L. Urine ketones of 4+ (160 mmol/L) were found on dipstick when blood beta-hydroxybutyrate levels exceeded 2 mmol/L. Seizure control correlates with blood beta-hydroxybutyrate levels and is more likely when blood beta-hydroxybutyrate levels are greater than 4 mmo/L. The traditional measurement of urine ketones by dipsticks in children on the ketogenic diet provides a less than optimal assessment of the degree of blood ketosis. Three to four plus (80-160 mmol/L) urine ketones are necessary, but not necessarily sufficient, to achieve optimal seizure control in children on the ketogenic diet. At present, however, urine ketones are the only readily available inexpensive approach to ketone assessment.