1. Intracellular and extracellular recordings were obtained from neocortical brain slices of immature rats (postnatal days 9-16) maintained in vitro. Spontaneous and evoked epileptiform discharges (termed paroxysmal depolarizing shifts or PDSs) were recorded from upper cortical laminae (layers II-III) after exposure to the γ-aminobuturic acid-A receptor antagonist, bicuculline methiodide. The effects of mGluR activation on PDS duration, spontaneous frequency, and threshold for evoking a PDS were determined. Putative mGluR agonists and antagonists also were tested. 2. Bath application of the mGluR agonist (1S,3R)-1-aminocyclopentane-1,3- dicarboxylate (ACPD, 50-200 mM) elicited biphasic, time-dependent effects on evoked and spontaneous epileptiform discharges. At times early in drug wash- in, ACPD increased PDS duration and spontaneous PDS frequency. In >60% of the slices, the spontaneous PDSs became regular. Subsequently, ACPD reduced PDS duration and increased the stimulus threshold for evoking a PDS, suggesting that the actions of ACPD were dose dependent. 3. Investigation of the concentration-dependence revealed that sustained low ACPD concentrations (5 μM) elicited only facilitatory actions, whereas higher concentrations were suppressive. These observations suggest the activation of different mGluR subtypes, which may be localized differentially at pre- and postsynaptic sites. 4. Bath application of the mGluR agonists, 1-2-amino-4- phosphonobutyrate or (2S,3S,4S)-α-(carboxycyclopropyl)glycine, produced only suppressive effects on epileptiform activity in the immature neocortex. L-2- amino-3-phosphonopropionate was an ineffective antagonist of ACPD-mediated modulation of epileptiform activity. Application of the putative antagonist, α-methyl-4-carboxyphenylglycine (MCPG), failed to antagonize the biphasic actions of ACPD. MCPG had suppressive effects of epileptiform activity, suggesting activation of mGluRs by endogenous agonists. 5. Simultaneous recordings from deeper and upper conical layers indicated that the initial negativity of both evoked and spontaneous PDSs began in deeper cortical layers under control conditions and in the presence of ACPD. Intracellular records from neurons in deeper layers displayed two distinct patterns of activity during mGluR activation. Most deep layer neurons received a barrage of excitatory postsynaptic potentials before a spontaneous PDS during ACPD application. A small population of neurons depolarized and entered a tonically firing mode, which was interrupted by spontaneous PDSs. Different neuronal populations, possibly expressing different mGluR subtypes or coupling mechanisms, may play integral roles in the induction and generation of epileptiform activities. 6. Thapsigargin or dantrolene, agents thought to block release of Ca2+ from intracellular stores, were bath applied for periods ≤35 min. Neither the facilitatory nor the suppressive effects of ACPD were antagonized by either drug. 7. These findings suggest that, during the early postnatal period, activation of neocortical mGluRs has significant facilitatory and suppressive effects on bicuculline-induced epileptiform activity. The facilitation may underlie in part the long-duration epileptiform events seen in immature neocortex. Expression of mGluR subtypes may differ between cortical laminae, neuronal subpopulations, and pre- versus postsynaptic locations. These differences appear to be regulated developmentally.