The highly mu-selective agonist Tyr-D-Ala-Gly-MePhe-Gly-ol-enkephalin (DAGO) produces potent, dose-dependent naloxone-reversible antinociception when microinjected into the ventrolateral periaqueductal gray (PAG) (ED50 = 0.72 nmol) or rostral ventromedial medulla (RVM) (ED50 = 0.05 nmol) as measured on the rat tail flick (TF) assay. In single-unit recording experiments, DAGO microinjected into the PAG also affected On- and Off-Cell firing in the RVM in the same way as previously demonstrated by our group for morphine. PAG-microinjected DAGO inhibits spontaneous and noxious-evoked On-Cell firing (attenuating the characteristic On-Cell burst) (n = 19), and excites spontaneous Off-Cell firing, preventing the characteristic Off-Cell pause (n = 12) at doses which suppress the TF. These results support a major role for the mu receptor in PAG and RVM mechanisms of opiate antinociception. In our experiments using BAM22P, an endogenous weakly mu-selective opioid peptide, we could not demonstrate a dose-dependent antinociceptive effect, whether the peptide was microinjected supraspinally into the PAG (n = 9) or RVM (n = 11), or intrathecally at the lumbar cord (n = 4). In two animals, a naloxone-reversible antinociceptive effect was observed following the microinjection of 10 nmol BAM 22P into the RVM; however, no effect was seen in 3 animals microinjected with 20 nmol. Dyn A(1-13), a putative endogenous ligand for the kappa receptor, had no antinociceptive effect when microinjected into the ventrolateral PAG, and no effect on the firing (spontaneous or noxious-evoked) of RVM On (n = 3)- or Off (n = 2)-Cells.