1. The [3H]-D-aspartate preloading of the parietal cortex of freely moving guinea-pigs equipped with epidural cups makes it possible to investigate drug effects on the efflux of this radiolabel, assumed as a marker of the glutamatergic structures underlying the cup. In the same model, the efflux of [3H]-gamma-aminobutyric acid ([3H]-GABA) and endogenous GABA can be measured. 2. Nicotine, 0.9-3.6 mg kg-1, s.c., or 3-5 micrograms, i.c.v., increased the efflux of [3H]-D-aspartate but reduced that of GABA. 3. These effects were mediated through mecamylamine-sensitive receptors but the ganglionic blocking agent was devoid of any primary activity. 4. The inhibition of GABA efflux induced by nicotine 3.6 mg kg-1, s.c., was abolished by methysergide 2 mg kg-1, i.p. and was reduced by naloxone 3 mg kg-1, i.p. pretreatment, suggesting the involvement of tryptaminergic and opioid systems. In contrast, muscarinic and catecholamine antagonists were ineffective. 5. Chronic treatment with nicotine (3.6 mg kg-1, twice daily for 16 days) reduced the facilitatory effect of [3H]-D-aspartate and abolished the inhibition of endogenous GABA efflux. 6. A slight increase in the number of nicotinic binding sites (by use of [3H]-nicotine as ligand) was found in the neocortex of chronically treated guinea-pigs. 7. The higher degree of tolerance to chronic nicotine treatment shown by GABA as compared with [3H]-D-aspartate efflux suggests that adaptative changes of the inhibitory neuronal pools prevail. This may contribute to the reinforcing and addictive properties of nicotine.