We recently demonstrated, in rat brain slices, that the usual excitation by noradrenaline (NA) of hypocretin/orexin (hcrt/orx) neurons was changed to an inhibition following sleep deprivation (SD). Here we describe that in control condition (CC), i.e. following 2 hours of natural sleep in the morning, the α(2)-adrenergic receptor (α(2)-AR) agonist, clonidine, had no effect on hcrt/orx neurons, whereas following 2 hours of SD (SDC), it hyperpolarized the neurons by activating G-protein-gated inwardly rectifying potassium (GIRK) channels. Since concentrations of clonidine up to a thousand times (100 µM) higher than those effective in SDC (100 nM), were completely ineffective in CC, a change in the availability of G-proteins is unlikely to explain the difference between the two conditions. To test whether the absence of effect of clonidine in CC could be due to a down-regulation of GIRK channels, we applied baclofen, a GABA(B) agonist known to also activate GIRK channels, and found that it hyperpolarized hcrt/orx neurons in that condition. Moreover, baclofen occluded the response to clonidine in SDC, indicating that absence of effect of clonidine in CC could not be attributed to down-regulation of GIRK channels. We finally tested whether α(2)-ARs were still available at the membrane in CC and found that clonidine could reduce calcium currents, indicating that α(2)-ARs associated with calcium channels remain available in that condition. Taken together, these results suggest that a pool of α(2)-ARs associated with GIRK channels is normally down-regulated (or desensitized) in hcrt/orx neurons to only become available for their inhibition following sleep deprivation.