The role of metabotropic (mGluRs) and N-methyl-D-aspartate (NMDA) glutamate receptors on 5-hydroxytryptamine (5-HT) release has been studied in rat periaqueductal gray (PAG) matter by using in vivo microdialysis. (1S,3R)-aminocyclopentane-1,3-dicarboxylic acid [(1S,3R)-ACPD; 0.5 or 1 mM], a group I/group II mGluRs agonist, increased the dialysate 5-HT concentration. (2S)-alpha-ethylglutamic acid (EGlu; 1 mM), an antagonist of group II mGluRs, but not (RS)-1-aminoindan-1,5-dicarboxylic acid (AIDA; 1 mM), an antagonist of group I mGluRs, antagonized the 1S,3R-ACPD-induced effect. (S)-3,5-dihydroxyphenylglycine (DHPG; 0.5 and 1 mM), an agonist of group I mGluRs, did not modify dialysate 5-HT. (2S, 3S, 4S)-alpha-(carboxycyclopropyl)-glycine (L-CCG-I; 0.5 and 1 mM), an agonist of group II mGluRs, increased extracellular 5-HT. This effect was antagonized by EGlu. Similarly, L-serine-O-phosphate (L-SOP; 1 and 10 mM), an agonist of group III mGluRs, increased extracellular 5-HT and this effect was antagonized by (RS)-alpha-methylserine O-phosphate (M-SOP; 1 mM), an antagonist of group III mGluRs. Out of the several N-methyl-D-aspartate concentrations used (NMDA; 10, 50, 100, 500 and 1000 mu M) only the 50 mu M infusion significantly decreased dialysate 5-HT. The GABA, receptor agonist, bicuculline (30 mu M), increased 5-HT release on its own and antagonized the decrease caused by the opiate antagonist, naloxone (2 mM), as well as the increases caused by CCG-I or L-SOP. These data show that stimulation of PAG's group II/group III mGluRs increases 5-HT release, while stimulation of NMDA glutamate receptors may decrease it. We speculate that glutamate does not modulate 5-HT release in the FAG directly, but via activation of tonically active GABAergic interneurons.