From the following three lines of evidence, it is proposed that at least part of the convulsant activity of naloxone is a result of GABA receptor blockade. Firstly, iontophoretic naloxone reversibly antagonized GABA-evoked depression of firing rate in 21 of 27 neurons tested in the rat olfactory tubercle-nucleus accumbens region, without blocking inhibition evoked in the same cells by glycine (15 cells) or morphine (6 cells). Secondly, i.p. naloxone in high doses caused convulsions in mice, and potentiated the convulsant activity of bicuculline, but not that of strychnine. Diazepam, which protected mice against convulsions elicited by bicuculline, but not by strychnine, also protected mice against naloxone. Thirdly, naloxone, morphine, levorphanol and its non-analgesic enantiomer dextrorphan displaced 3H-GABA from GABA receptor sites in homogenates of human cerebellum, all with comparable low potencies (IC50 = 250--400 micron). There was no correlation with affinities at the stereospecific receptor sites that mediate opiate-induced analgesia, since the potent opiates etorphine and diprenorphine were relatively inactive (IC50 greater than 3 mM). In addition naloxone displaced 3H-GABA from receptor sites in rate forebrain and cerebellum, with similar low potency.