The effects of helium pressure and of general anaesthetics were studied on the responses of the isolated superior cervical ganglion of the rat, to determine how far these reflected the pressure reversal of anaesthesia seen in vivo. The method of Brown & Marsh (1974) for extracellular recording of surface potentials was adapted for use in a high-pressure chamber. Helium alone, at 130 atm, did not alter the responses of the ganglion to gamma-aminobutyric acid (GABA) but significantly depressed the depolarizing and hyperpolarizing components of the nicotinic responses, and the muscarinic responses. The potentiation of the responses to GABA caused by pentobarbitone was not altered by the application of helium, at 130 atm. This pressure also decreased further the nicotinic responses which were depressed by pentobarbitone. Nitrogen, at 34 atm (the anaesthetic ED50 in vivo) and at 68 atm, significantly decreased the nicotinic responses of the ganglia, and the addition of helium to a total of 130 atm further increased this depression. At pressures of 3.3-68 atm, nitrogen caused small decreases in the responses to GABA. Nitrous oxide at 1.5 atm (the ED50 for loss of righting reflex in mice) and at 3 atm, significantly depressed the responses to GABA and to the nicotinic agonist, but did not alter the responses to methylfurmethide. The effects of nitrous oxide were unaltered when helium was added to a total of 130 atm, although this pressure of helium added alone significantly depressed the cholinergic responses. A mixture of 50% nitrous oxide and 50% oxygen, when added to the pressure chamber, at normal atmospheric pressure, caused transient increases in the responses to GABA. The effects of temperature on GABA responses and on nicotinic responses were very different from those of pressure. Preliminary evidence suggested that raising the temperature may decrease the extent of potentiation of GABA responses by pentobarbitone. The results are discussed in relation to the pressure reversal of anaesthesia in vivo. It was concluded that there was no evidence that the basis of this interaction lay in the potentiation of GABA responses by general anaesthetics, or the depression of cholinergic responses, although the changes seen were not in all cases simply additive. It was considered that effects of general anaesthetics such as the potentiation of GABA may contribute to the effects used to measure general anaesthesia in vivo, such as loss of righting reflex, but may not be related to the non-specific actions which cause anaesthesia.