Ninety-six female introductory psychology students differing in terms of the dimensions of neuroticism-stability and introversion-extraversion performed two sets of thirty trials each on a choice reaction time task. On one of the sets of trials the Ss were observed by two female observers and on the other set of trials the observers were not present. The order of observer presence was counterbalanced in each of the four groups (neurotic introverts, neurotic extraverts, stable introverts and stable extraverts) of 23 Ss. The mean reaction time score of each S under the audience and the no audience conditions was analyzed using a 2 (neuroticism-stability) x 2 (extraversion-introversion) x 2 (order of audience conditions) mixed analysis of variance. The Extraversion x Neuroticism x Audience condition interaction was found to be significant at the .005 level of confidence. This interaction indicated the presence of observers inhibited the performance of neurotic introverts and facilitated the performance of stable extraverts. The Extraversion x Order x Audience condition interaction was found to be significant at the .001 level of confidence. This interaction indicated that the presence of others facilitated the performance of extraverts only if the audience condition preceded the no audience condition. (Author)