Recent evidence indicates that the basolateral amygdala (BLA) may be involved in behavioural effects induced by cannabinoids. High levels of CB1 cannabinoid receptors have been shown in this region, where they modulate excitatory and inhibitory synaptic transmission. However, the neurophysiological effects of these opposing synaptic actions have not been investigated in vivo. To this purpose, single-unit extracellular recordings were performed in urethane anaesthetized rats in order to determine whether exogenously applied cannabinoids influenced the spontaneous or evoked electrical activity of neurons in the BLA. The effects of cannabinoids were found to be dependent on the characteristics of the neurons examined and on the properties of the agents used. We tested and compared two structurally different synthetic cannabinoid receptor agonists, the highly potent HU-210 (0.125-1.0 mg/kg, i.v.) and WIN55212-2 (WIN, 0.125-1.0 mg/kg, i.v.). With a CB1 cannabinoid receptor-dependent mechanism, HU-210 potently inhibited the firing rate of BLA interneurons whereas WIN modulated the discharge rate in a biphasic manner. By contrast, BLA projection neurons, antidromically identified from the shell of the nucleus accumbens, were significantly inhibited by WIN at all doses tested, while HU-210 administration led to less consistent effects, since only 1.0 mg/kg inhibited firing rate in the majority of recorded neurons. Additionally, WIN, but not HU-210, significantly attenuated short-latency spiking activity in BLA projection neurons evoked by electrical stimulation of the medial prefrontal cortex. In these neurons, WIN-induced effects were antagonised by the non-selective cannabinoid receptor antagonist SR141716A and by the vanilloid receptor antagonist capsazepine, but not by the selective CB1 antagonist AM-251. Taken together, our findings indicate that the overall excitability of efferent neurons in the BLA is strongly reduced by WIN in a non-CB1-dependent manner. In this effect, the contribution of a novel cannabinoid-vanilloid-sensitive putative non-CB1 receptors, the existence of which was postulated in recent reports, might play a role.