Slices of rabbit cerebral cortex were preincubated with [3H]-noradrenaline and then superfused and stimulated electrically twice for 2 min each (S1, S2) at various frequencies (0.2-3 Hz). The stimulation-evoked overflow of tritium (S1) increased with increasing frequency and was higher when cocaine (10 microM) was present. In the absence of cocaine, tetraethylammonium (TEA; 100 and 300 microM), added before S2, increased the stimulation-evoked overflow of tritium to about the same extent, irrespective of the frequency. In contrast, rauwolscine (0.1 and 1 microM) and idazoxan (0.1-10 microM) increased the evoked overflow much more, the higher the frequency of stimulation. Phentolamine (0.1 and 1 microM) reduced the overflow elicited at 0.3 and 1 Hz, and (1 microM) caused an increase only at 3 Hz. In slices superfused throughout with cocaine 10 microM, rauwolscine (1 microM) and idazoxan (1 and 10 microM) again increased the evoked overflow of tritium more, the higher the frequency of stimulation. For a given frequency, rauwolscine and idazoxan enhanced the evoked overflow to a greater extent in the presence than in the absence of cocaine. Idazoxan (1 and 10 microM) and rauwolscine (1 microM) counteracted the inhibition that phentolamine (0.1 microM) produced at low frequency. The increases caused by rauwolscine (1 microM) and TEA (300 microM) were approximately additive, but those caused by rauwolscine (1 microM) and idazoxan (10 microM) were not. The effects of rauwolscine, idazoxan and phentolamine depend on the experimental conditions (frequency, cocaine) in a manner compatible with the operation of a presynaptic alpha 2-adrenoceptor-mediated autoinhibition of noradrenaline release. When given at sufficient concentrations, these antagonists enhance the release of noradrenaline more, the higher the biophase concentration of the transmitter and the stronger, hence, the autoinhibition. In the case of the partial alpha 2-adrenoceptor agonist phentolamine, a low perineuronal noradrenaline concentration even reverses facilitation to inhibition. This pattern differs markedly from the pattern of effects of TEA which increases the release of noradrenaline by a mechanism other than alpha-adrenoceptor blockade.