Behavioral, endocrinological, and pharmacological data suggest that the emotional response of rodents to the elevated plus-maze alters as a function of prior test experience. In the present study, 74 intact male Swiss-Webster mice were exposed to the plus-maze for 5 min on each of 3 consecutive days, with all test sessions recorded on videotape. Behavior patterns for each trial were scored using ethological analysis software and the resultant database subjected to a number of statistical treatments. Analysis of full session profiles (i.e., 5 min total scores) showed that a single prior undrugged experience of the maze increases behavioral indices of anxiety and that these alterations are either maintained or further enhanced on subsequent trials. Furthermore, the behavioral profile evident by trial 3 was largely unchanged when animals were reexposed to the maze 10 days later. More detailed (i.e., min by min) examination of behavior patterns within and between trials demonstrated that unambiguous open arm avoidance is acquired by the third minute of trial 1, and that the behavioral profile evident by the end of trial 1 is (a) markedly different to that seen at the beginning of that trial, and (b) generally maintained or even accentuated on trials 2 and 3. The implied impact of prior test experience on future behavioral strategy in the maze was strongly supported by a series of factor analyses. Thus, while the factor associations of vertical activity and directed exploration remained constant across trials, trial 2 and 3 anxiety measures loaded on a separate factor to that loading trial 1 anxiety measures. A similar trial 1 vs. trials 2 and 3 dissociation was observed for measures of locomotor activity. Although the present findings are consistent with the proposal that prior test experience produces a qualitative shift in emotional response to the elevated plus-maze, the precise basis for this change as well as its full significance for our understanding of anxiety-related processes remain to be determined.