Brain slices have been used for several years in the study of the factors controlling neurotransmitter release in animals. Despite this, most conventional studies have drawbacks. First, it is often necessary to label the neurotransmitter pool with tritium (Farnebo and Hamberger, 1971). This assumes, probably fallaciously, that release of the radiolabel accurately reflects the neurotransmitter. For example, when endogenous and radiolabeled catecholamine releases are simultaneously measured in response to various stimuli, there are significant discrepancies between the two (Herdon and Nahorski, 1987). Second, transmitter release is frequently evoked by elevating the potassium concentration. However, some transmitter is released in a calcium-independent fashion following potassium depolarization (Okuma and Osumi, 1986). Third, many superfusion methods need to block transmitter uptake in order for efflux to be measurable (Parker and Cubeddu, 1985). These are serious limitations.