1. The effect of pentobarbitone (PB) on GABA-ergic inhibition was investigated in the isolated crayfish stretch receptor. The soma of the slowly adapting neurone was impaled with two micro-electrodes to give an accurate determination of membrane conductances. 2. Application of PB in concentrations from 10(-6) to 10(-3) M increased the rise time constant of the inhibitory post-synaptic potential (i.p.s.p.). The i.p.s.p. percentage amplitude and decay time constant were also increased in eight out of twelve neurones. On prolonged exposure, the percentage amplitude declined at a rate dependent upon the dose and the frequency of stimulation until the i.p.s.p. became undetectable. 3. The response to ionophoretically applied GABA remained essentially unaltered in the presence of PB, but the falling phase was prolonged by up to 8% in four of the ten neurones tested. Resting membrane conductance, i.p.s.p. driving force (i.p.s.p. reversal potential minus resting membrane potential), and parameters of the anti- and orthodromic action potential were not significantly affected. 4. Removal of PB after prolonged exposure usually caused an immediate increase in i.p.s.p. percentage amplitude but the i.p.s.p. rising phase remained slowed. 5. Application of excess extracellular GABA only affected the i.p.s.p. percentage amplitude after it had been reduced by PB. It transiently increased the attenuated i.p.s.p. percentage amplitude in the presence of PB, and after the removal of PB permanently increased the amplitude to its original value. 6. Nipecotic acid and cis-1,3-aminocyclohexane carboxylic acid, inhibitors of GABA re-uptake, slightly increased the i.p.s.p. percentage amplitude, and prolonged the falling phase but did not affect the rising phase. The percentage amplitude declined on prolonged exposure. 7. We conclude that PB has no electrophysiologically demonstrable post-synaptic action in the crayfish stretch receptor neurone, but it inhibits the presynaptic release and re-uptake of GABA.