Effect of three doses of rimonabant on improgan antinociception in the PAG. Two consecutive drug injections were made through a single cannula located in either the dorsal or ventrolateral PAG (see Fig. 1). Experiments were identical with those described in Fig. 2, except for cannula locations, and all subjects received improgan. Pre-treatment consisted of either vehicle (Veh) or rimonabant (5, 20, or 40 μg, SR5, SR20, SR40, respectively). Data from DPAG and VLPAG experiments showed no significant differences, and were pooled. Results are shown in the format of Fig. 2. *,**P < 0.05, 0.01 vs. Veh/Imp at the same time.