Writers from a number of theoretical backgrounds have asserted that agreement in the emotional messages conveyed by various verbal and nonverbal communication channels is related to the communicator's psychological health. If this conjecture is accurate, then congruence among communication channels could be used as a behaviorally based assessment tool. However, empirical research to test this theoretical and clinical assumption is relatively lacking. The present study was designed to test the hypothesis that individuals who display congruence (agreement) between verbal (language), verbal/vocal (language plus paralinguistic cues, or speech) and nonverbal (facial) channels of communication will show a greater degree of mental health than will individuals who display incongruence. "Degree of mental health" was operationally defined as an individual's scores on the Personal Orientation Inventory (POI). Fifty-six subjects were administrated the POI and were interviewed on videotape. Three pairs of judges rated the videotapes for the affects communicated in the video channel (picture only), the audio channel (sound only), and the transcript channel (the subject's words transcribed onto paper). Comparisons of affect ratings across channels yielded difference scores, resulting in measures of various types of congruence. Analyses of variance were carried out with difference scores as independent variables and and overall POI score as the dependent variable. No significant results were obtained. Multivariate analyses of the POI subscales were also performed, again with nonsignificant findings. Alternative explanations of the congruence phenomenon and methodological limitations are presented. Implications for the clinical utility of congruence and for future research are discussed.