Louis-Sylvestre and LeMagnen have suggested that the premeal decline in blood glucose is or reflects a signal for meal initiation in rats. In order to extend and test this hypothesis, a computer controlled system for continuously and concurrently measuring blood glucose and food intake in free-feeding rats was developed. In 18 experiments (with and without intravenous saline infusions), blood glucose declined about 12 minutes prior to meal onset. During 2-1/2 hours of observation, no decline in blood glucose and no meal occurred in 19 other experiments. In 7 experiments in which 10 percent glucose was infused IV to partially block the premeal decline (average blockade = 46.5%), the subsequent meal was significantly delayed. These results suggest that the pre-meal decline in blood glucose is not only correlated with but is also caudally related to meal onset. These studies suggest that the premeal decline in blood glucose is among the signals for meal initiation.