Human immunodeficiency virus 1 (HIV-1)-associated dementia (HAD) represents a common complication of HIV-1 infection. Antiretroviral therapy has diminished its incidence, but it is insufficient to eradicate the problem. HAD depends on the presence of the virus in central nervous system (CNS), but the molecular mechanisms involved are not completely understood. It is widely accepted that proteins shed by the virus, such as the envelope glycoprotein gp120 and the nonstructural viral protein Tat, may themselves cause alterations to CNS. By one side, viral proteins are toxic to neurons because of their ability (1) to act as excitotoxins and (2) to evoke the release of endogenous neurotoxins and/or proinflammatory cytokines. By the other side, evidences are emerging that viral components can alter neuronal functions either by modifying the release of neurotransmitters or by influencing the functions of classical receptors controlling central neurotransmission. We here review some results concerning the effects of Tat on cholinergic and noradrenergic neurotransmission in human and rat cortex. The protein can induce the release of acetylcholine from both human and rat cortical cholinergic nerve terminals in a specie-specific manner. In human cholinergic terminals, Tat-mediated releasing effect depends on activation of receptors belonging to I group of metabotropic glutamate receptors (mGluRs), while in rat terminals Tat-induced effect involves the activation of a so far unknown receptor. The protein, unable on its own to release noradrenaline from human and rat cortical noradrenergic nerve endings, potentiates the release of amine induced by presynaptic NMDA receptors. Also in this case, Tat effect involves activation of a receptor belonging to the group I mGluRs, in particular of the mGluR1 subtype. The finding that group I mGluRs may represent a preferential target of the protein in CNS may be relevant to the proposal of new therapeutic approaches for the cure of HAD.