Fast habituation (FH) is defined as a decrease in auditory evoked potential (AEP) amplitude in response to the second of a pair of temporally juxtaposed (e.g. 1-s) tones. The degree of FH may depend, in part, on the subject's ability to anticipate the stimulus sequence. This paradigm has been used in our laboratory to study cognitive functioning in human subjects. We have also developed an animal model to investigate, more comprehensively, the anatomical and physiological basis of this phenomenon seen in human subjects. In the present investigation, we wished to determine the relationship between single-trial latency variability and AEP amplitude for two conditions: one for which FH is known to occur and one for which it is not present, due to the length of the interstimulus interval (ISI). Here AEPs were obtained to 40 pairs of 100-ms pure tone stimuli (70-dB SPL, 1- and 5-s ISI, 10-s interpair interval) from 19 chronically implanted, unanesthetized, restrained male Sprague-Dawley rats. The AEP latencies, amplitudes, and the single-trial latency variabilities were obtained for each component (P1, N1, P2 and N2). Findings indicated that FH was present for the 1-s ISI condition but not for the 5-s ISI condition. In addition, single-trial latency variability was negatively correlated with both Tones 1 and 2 AEP amplitudes in the 5-s ISI condition but only with Tone 1 amplitude in the 1-s ISI condition. Thus, single-trial latency variability predicted AEP amplitude only when FH did not occur. These results support earlier findings reported from our laboratory suggesting that the decrease in amplitude during FH is not related to increased variability in the time domain but rather to decreased neuronal output.