The various effects of temperature change are only partially predictable. Temporal measures relevant to membrane activity, action potentials, synaptic transmission, and evoked potentials are all consistently increased with cooling and decreased by warming. However, the various measures of amplitude at different levels, and even within similar preparations, are contradictory: Some laboratories report increased amplitudes with cooling and others report decreased amplitudes under similar conditions. Emphasis is given to identifying factors that may resolve the differences. These include: (a) the rate of temperature change, (b) sites of cooling, stimulation and recording, (c) stimulus characteristics, and (d) fundamental differences in temperature sensitivities of different neural tissue. Other factors that may affect the ability to predict thermal influences on neural function from existing formulations are: relative ion permeabilities, metabolic ion pumps, the resting potential at the onset of cooling, and an animal's acclimated temperature at sacrifice.