The purpose of this study is to clarify the temporal structure of utterance and looking patterns in a face-to-face dialogue, considering its relation with the presence of anxiety discrepancy between dyad speakers.The subjects are sixty female freshmen in two colleges. Thirty pairs of female students are all grouped into six MAS combinations of high, middle and low anxiety. Subjects, not acquainted previously, are asked to talk about one of the moderate interesting topics in two 24-minute sessions one or two-days apart. Communication data in the first session are chronographically analyzed by some interaction indices.In this study four verbal activity indices are employed as zero-state temporal sequence -Common Silence, One's Only one Talking, Partner's Only one Talking, and Common Talkingand several indices as one- or two-state temporal sequence.Four looking activity indices are also employed as zero-state temporal sequence-Neither Looking, One's Only one Looking, Partner's Only one Looking, and Both Looking-and one index as one-state sequence.Principal factor analysis was applied to these 50 indices. As the result it proves that verbal activity factors are independent of looking activity factors. Main factors in this analysis are the general cooperative activity and the individual activity in each verbal and looking activity. In addition to these factors, the following main factors are obtained: partner's utterance, partner's looking, pausing in conversation, one's utterance after silence, partner's utterance after silence, intensity of silence and partner's duration of utterance.The structure of verbal activity factors in this analysis is similar to the factors in non-faceto-face conversational research.There were found differences of the contribution level of general cooperative activity factor in each conversational group. This means that the presence of discrepancy of anxiety levels in two speakers tend to activate the general cooperative verbal activity and reduce the general cooperative looking activity Concerning individual activity, it is shown that the interrelation of utterance is MAS non-discrepant group>MAS discrepant group, but, this relationship is reversed as to individual looking activity level. Therefore, it may safely be said that there exist dual complementary relationships between general cooperative activity and individual activity with regard to two aspects such as talking and looking. While the vocal channel is independent of the visual channel, these communication channels have complicated inter relations with each other under the condition of interactants' relationship.Among activity factors which separate the anxiety-discrepant dyads from the anxiety nondiscrepant dyads, two factors-pausing in conversation, and cooperative verbal activity factors -are more discriminant than looking activity factors. As to individual activity, the more a speaker talks, the less a partner does for keeping an optimal degree of fluent dyadic conversation.Furthermore, indices of looking and utterance have slightly positive correlations in speaker-partner dyad.For interpersonal communication study, we need to take a multichannel approach, not an additive simple view. It is necessary to consider the meaning of interactants' relationship in conversation as an important variable.