This study was undertaken to examine whether the relationships reported in the communication literature between Eysenck's BIG THREE personality dimensions and self-reported patterns of communication behavior remain robust when observation-based measures are used in place of self-reports. Target participants' responses to measures of these three personality dimensions were used to predict peer ratings of two patterns of communication behavior, verbal aggressiveness and affinity-seeking competence. Composite measures for variables were based on the results of confirmatory factor analysis, ensuring both internal and external consistency (parallelism). Results of a path analysis, in which targets' responses to personality measures were employed as exogenous variables and peer ratings of verbal aggressiveness and affinity-seeking competence were used as endogenous variables indicated that these data fit the model extremely well. Importantly, (1) substantial variation in peer rated affinity-seeking competence (R=.55) was attributed to the direct effects of extraversion and neuroticism (negative coefficient), and (2) substantial variation of peer rated verbal aggressiveness (R=.58) was attributable to direct effects of psychoticism and extraversion (negative coefficient) and an indirect effect of extraversion and neuroticism mediated by affinity-seeking competence (negative coefficient). Implications for communication theory and research are discussed.