Cocaine (COC) has been described as exerting potent stimulatory effects on the hypothalamo-pituitary-adrenocortical (HPA) axis. In the present study, we investigated the acute and chronic effects of intracerebroventricular and intrahypothalamic injections of COC in rats. Twenty minutes following intracerebroventricular injection of COC (1-100 µg), dose-dependent increases in plasma corticosterone (CS) were observed, although the highest dose tested (100 µg) evoked a significantly smaller response than that following 50 µg. Prior stressing of the animals resulted in elevated plasma CS levels (315 ± 16 ng/ml) and significantly decreased plasma CS concentrations following 50 µg COC (87.8 ± 3.2%). Injections above the hypothalamic paraventricular nucleus (PVN), the site of corticotropin-releasing-factor-secreting neurons which regulate HPA activity, required relatively higher doses of COC in order to elicit increases in plasma CS; injections of 0.5 µg had no effect, 1 µg resulted in an increase to 168 ± 68 ng/ml (p < 0.005), and 2.5 µg produced an increase to 146 ± 29 ng/ml (p < 0.025). Post-PVN injections of COC, behind the posterior margin of the PVN in the vicinity of the ventral noradrenergic ascending bundle, also required a high dose (2.5 µg) in order to elicit a plasma CS response (208 ± 19 ng/ml; p < 0.005), with no significant response seen following 0.5 µg COC. No effects of specific neurotoxic lesions of the catecholaminergic or serotonergic innervation of the hypothalamus were observed upon adrenocortical responses to COC. Chronic intracerebroventricular administration of COC (5 µg/day, 10 days) to conscious rats resulted in significant elevations in basal plasma CS (238 ± 36 ng/ml; p < 0.025) above that found in saline-treated animals (129 ± 19 ng/ ml). However, animals receiving daily COC injections were able to exhibit a similar CS response (336 ± 26 ng/ml) to that seen in the chronic saline-treated group given an acute intracerebroventricular COC challenge (328 ± 23 ng/ml). Neurochemical changes in monoamine metabolism were also measured in medial prefrontal cortex, hippocampus, and hypothalamus following single or repeated daily injections of COC, and these are discussed in the context of the data on adrenocortical responses. The results of this study demonstrate that while a local intrahypothalamic action of COC may elicit HPA activation, probably through a stimulation of catecholaminergic activity, the predominant site of action of COC in such activation appears to reside elsewhere in the central nervous system. The elevated plasma CS concentrations observed even in saline-challenged rats following chronic COC administration may reflect an anxiety state similar to that seen in chronic human COC users. At the same time, however, these changes do not appear to alter the responsiveness of the HPA axis to subsequent COC administration.