In this study we examined whether the potency of quinolinic acid (Quin) in inducing neurodegeneration in vivo was dependent on the exposure time of the tissue to the excitotoxin. The effect of chronic infusion of Quin into rat striatum and hippocampus was examined at the light microscopic level and by cell count on 40 μm Cresyl violet stained brain sections. Continuous infusion was at a constant speed (0.5 μl/h) for various times (15 h–2 weeks) by osmotic minipumps (Alzet 2002). No build up of [3H]Quin occurred in the tissue during infusion; this was assessed by measuring the radioactivity 3–14 days after minipump placement. Intrastriatal infusion of 6 and 10 nmol/h Quin, but not of nicotinic acid, for 1 week induced a dose-dependent neurodegeneration (70 and 90% loss of neurons, respectively, compared to the contralateral striatum) extending 1.2–2 mm from the centre of the injection. The onset of the neurotoxicity caused by 10 nmol/h Quin was >24 h. One week's infusion of 4 nmol/h Quin did not induce neurotoxicity, but a 40% drop of neurons, compared to the contralateral side, occurred after 2 weeks. One week's intrahippocampal infusion of 2.4 and 6 nmol/h Quin, but not of nicotinic acid, caused a dose-dependent neurodegeneration with a radius of ∼1–1.5 mm around the injection track. The onset of the neurotoxicity induced by 2.4 nmol/h Quin was <15 h. The pattern of nerve cell loss induced by 1.2 nmol/h Quin after 1 week (CA4 cells lost in 50% of the rats) did not differ from that observed after 2 weeks of infusion. Nerve cell loss caused by Quin in the striatum and in the hippocampus was restricted to the injected area and antagonized by coinfusion with d(–)-2-amino-7-phosphonoheptanoic and kynurenic acids in molar ratios of 1:0.1 and 1:3, respectively. These data show that Quin's potency in inducing neurodegeneration in the striatum, but not in the hippocampus, depends on the exposure time of the tissue to the excitotoxin. In addition, neurodegeneration is induced faster by Quin in the hippocampus than in the striatum. The usefulness of this model to study the sequelae of the neurotoxic process in vivo will be discussed.