The Bales system of scoring small group interactions was used to analyze a transcript of a preamniocentesis (genetic) counseling session in which a husband (H), wife (W), and a male counselor (C) participated. Scoring required the subdivision of communication into the smallest meaningful and scorable units, assigning each unit one of twelve possible category scores and designating the initiator and intended recipient(s) of the unit. The categories included "positive" and "negative" reactions, questions and attempted answers. Over half the units were found to consist of interactions in which information, orientation, and clarification were provided. C initiated 54% of all units of which three-quarters consisted of factual information regarding amniocentesis, genetic risks, Down syndrome, and mental retardation. C interacted significantly more frequently with W than with H, but units directed toward H were significantly more positive than those toward W. One interaction, in which C responded to affective material, significantly altered the subsequent frequency and direction of units; H became more verbal and there was an increase in further affective expression. Although evidence of persuasion was present, no directiveness was demonstrated. The data show that quantitative study of genetic counseling is feasible. Further research on different counseling styles and on the influence of specific interventions on later outcome has been undertaken.