GABAA receptor positive allosteric modulators (PAMs) mediate robust analgesia in animal models of pathological pain. Restoration of diminished spinal GABAA-α2 and -α3 subunit-containing receptor function is a principal contributor to this analgesia, albeit involvement of GABAA-α5-receptors has not been excluded. Thus, we compared NS11394 and TPA023 (PAMs with selectivity/efficacy at GABAA-α2/α3/α5 receptors) with TP003 (a reportedly GABAA-α3 selective PAM) against spinal sensitization. However, in-house electrophysiology studies designed to confirm the selectivity of TPA023 and TP003 for human GABAA receptors did not corroborate published data, with TP003 displaying considerable GABAA-α5 receptor efficacy. Therefore, we identified a novel PAM, NS16085, which possesses negligible efficacy at GABAA-α5 receptors, but with GABAA-α2/α3 efficacy equivalent to NS11394. At the GABAA-α1 receptor the compound gives low level of negative modulation further separating it from the other compounds. Rat pups with carrageenan-induced hindpaw inflammatory hyperalgesia were used to make ex vivo spinal dorsal root-evoked ventral root recordings. Some spontaneous activity and large numbers of spikes to repetitive stimulation of dorsal roots at C-fibre intensity, indicative of wind-up and sensitization were observed. Equimolar concentrations of NS11394, TP003 and NS16085 all attenuated wind-up to a similar degree; TPA023 was clearly less effective. In adult rats, NS16085 (3-30 mg/kg, p.o.) dose-dependently reduced formalin-induced hindpaw flinching with efficacy comparable to NS11394. Thus, potentiation of GABAA-α2 and-α3 receptors is sufficient to depress spinal sensitization and mediate analgesia after inflammatory injury. Positive modulation at GABAA-α5-receptors is apparently dispensable for this process, an important consideration given the role of this receptor subtype in cognitive function.