The effects of transient hyperglycemia on brain glucose and the relationship between blood and brain glucose were studied in 76 Sprague-Dawley rats anesthetized with halothane 1% inspired. In a common control group, blood and brain glucose were determined prior to any intervention (n = 10). A second group of 30 rats was given an iv infusion of 3.9 ml of saline over a 30-min period, and blood and brain glucose were subsequently determined at several time points: 30 min (immediately after the saline infusion), 45, 60, 90, or 120 min (n = 6 for each time point). A third group of 36 rats was administered 3 g/kg of glucose in 3.9 ml of iv saline over a similar 30-min period, and blood and brain glucose were measured at the same time periods as in saline-treated rats and also when half of the glucose infusion was given (time = 15 min; n = 6 for each time point). In the common control group, blood glucose was 114 ± 14 mg/dl (6.4 ± 0.8 μmol/ml; mean ± SD) and brain glucose was 2.41 ± 0.59 μmol/g. Saline infusion had no effect on brain or blood glucose. In contrast, glucose infusion in the study group produced significant increases in both blood and brain glucose, achieving maximal values of 488 ± 60 mg/dl (27.4 ± 3.4 μmol/ml) and 7.62 ± 0.52 μmol/g, respectively, at the 30-min measurement period. The ratio of brain to blood glucose, normalized so that the common control group data achieved a value of 1, was less than unity during the period of glucose infusion. This ratio reached a nadir of 0.75 ± 0.07 at the 30-min measurement period (P< 0.05versussaline infusion). Thereafter, with the cessation of glucose infusion, the ratio returned to 1 and eventually exceeded unity: the peak ratios were 1.23 ± 0.13 and 1.16 ± 0.21 at the 60− and 90-min period, respectively (P< 0.05versussaline treatment). The authors concluded that during periods of rapidly fluctuating blood glucose, there is a hysteresis between blood and brain glucose values; hence, it may not be possible to accurately estimate brain glucose by measuring blood glucose.