Previously, we have described a selective potentiating effect of systemically administered cocaine (0.25-1.0 mg/kg i.v.) on long-latency excitatory responses (E2) of rat "barrel field" cortical neurons to mystacial vibrissae stimulation. The rat trigeminal system receives both norepinephrine (NE) and serotonin (5-HT)-containing afferents, but only minimal input from dopaminergic sources. The goal of the present study was to determine which of these monoamine systems was responsible for the previously observed facilitating action of cocaine on E2 responses of barrel field cortical neurons. Two approaches were used: 1) evaluation of cocaine effects on cortical neuron responses to whisker stimulation in NE- or 5-HT-depleted animals and 2) assessment of the effects of selective monoamine uptake blockers on cortical neuron responses to whisker deflection. Extracellular recordings were obtained from spontaneously active neurons in the barrel field cortex of halothane-anesthetized rats. Spontaneous activity and cellular responses to mechanical displacement of a single whisker were monitored before and after systemic (i.v.) administration of either cocaine or one of the following selective uptake blockers, fluoxetine (5-HT), desipramine (NE) and GBR12909 (dopamine). Cocaine-induced increases in the E2 response were observed in N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine (DSP4, noradrenergic neurotoxin)-treated animals, but were reduced or abolished in p-chlorophenylalanine-treated (5-HT depletion) rats. Fluoxetine and desipramine, but not GBR12909, produced cocaine-like potentiation of the E2 response to whisker stimulation. These results point to a 5-HT-dependent mechanism as the substrate underlying cocaine's facilitating effects on long-latency somatosensory cortical neuron responses to receptive field stimulation.