1 The role of a 2 -adrenoceptor (AR) subtypes in the modulation of acute nociception, motor behaviour and body temperature, has been investigated by determining the activity of the a 2 AR selective agonist dexmedetomidine (Dex) in mice devoid of individual a 2 AR subtypes through either a point (a 2A ) or null (a 2B /a 2C ) mutation (`knock-out').2 In a rodent model of acute thermal nociception, the mouse tail immersion test, Dex, in wild type (WT) control animals, produced a dose-dependent increase in the threshold for tail withdrawal from a 528C water bath with mean ED 50 values of 99.9+14.5 (a 2A ), 94.6+17.8 (a 2B ) and 116.0+17.1 (a 2C ) mg kg 71 , i.p. 3 In comparison to the WT controls, Dex (100 ± 1000 mg kg 71 , i.p.), was completely ine ective as an antinociceptive agent in the tail immersion test in the a 2A AR D79N mutant animals. Conversely, in the a 2B AR and a 2C AR knock-outs, Dex produced a dose-dependent antinociceptive e ect that was not signi®cantly di erent from that observed in WT controls, with ED 50 values of 85.9+15.0 (P40.05 vs WT control) and 226.0+62.7 (P40.05 vs WT control) mg kg 71 i.p., respectively. 4 Dex (10 ± 300 mg kg 71 , i.p.) produced a dose-dependent reduction in spontaneous locomotor activity in the a 2A , a 2B and a 2C AR WT control animals with ED 50 values of 30.1+9.0, 23.5+7.1 and 32.3+4.6 mg kg 71 , i.p., respectively. Again, Dex (100 ± 1000 mg kg 71 , i.p.) was ine ective at modulating motor behaviour in the a 2A AR D79N mutants. In the a 2B AR and a 2C AR knock-out mice, Dex produced a dose-dependent reduction in spontaneous locomotor activity with ED 50 values of 29.1+6.4 (P40.05 vs WT control) and 57.5+11.3 (P40.05 vs WT control) mg kg 71 , respectively. 5 Dex was also found to produce a dose-dependent reduction in body temperature in the a 2A , a 2B and a 2C AR WT control mice with ED 50 values of 60.6+11.0, 16.2+2.5 and 47.2+9.1 mg kg 71 , i.p., respectively. In the a 2A AR D79N mutants, Dex had no e ect on body temperature at a dose (100 mg kg 71 , i.p.) that produced a signi®cant reduction (76.2+0.58C; P50.01 vs vehicle) in temperature in WT controls. However, higher doses of Dex (300 and 1000 mg kg 71 , i.p) produced a small, but statistically signi®cant decrease in temperature corresponding to 71.7+0.48C and 72.4+0.38C (both P50.01 vs vehicle), respectively. In the a 2B AR and a 2C AR knock-out mice, Dex produced a dose-dependent reduction in body temperature with ED 50 values of 28.4+4.8 (P40.05 vs WT control) and 54.1+8.0 (P40.05 vs WT control) mg kg 71 , respectively. 6 In conclusion, the data are consistent with the a 2A AR being the predominant subtype involved in the mediation of the antinociceptive, sedative and hypothermic actions of Dex. This pro®le would appear to indicate that an a 2A AR subtype selective analgesic will have a narrow therapeutic window, particularly following systemic administration.