As with any neurotransmitter, there are two basic ways to decrease the synaptic action of γ-aminobutync acid (GABA) in order to establish its functional role at particular synapses in the nervous system. They are to reduce its availability and release through a presynaptic action or to reduce its effects with appropriate postsynaptic antagonists. The former might be achieved by specifically destroying GABA-containing neurons, inhibiting its synthesis, or interfering in some way with its actual release. All of these approaches, however, suffer from the disadvantage of a very slow onset and an effect that, even if reversible, is likely to be prolonged and not easily quantified. There are no specific neurotoxins for GABA neurons similar to 6-hydroxydopamine for catecholamme systems, and although antibodies raised to the GABA-synthesizing enzyme, glutamic acid decarboxylase (GAD), can be used to destroy (lyse) synaptosomes containing GABA in vitro (Docherty et al., 1983), the method is not applicable in vivo. Intracerebral injection of tetanus toxin has been shown to reduce GABA-mediated inhibition by reducing release of the transmitter (Collingridge and Davies, 1982), but this toxin has a number of actions, including an effect on glycine release, and it is obviously not specific for GABA. Many inhibitors of GAD have been described (see Woodbury, 1980); unfortunately, many of them not only reduce GABA levels, but also raise the concentration of its precursor glutamate. Their use also suffers from the disadvantage that the rate of decline of GABA depends on the rate of its utilization, which will vary from neuron to neuron, and we do not know to what extent GABA levels must be reduced before its release and function are significantly impaired.