Students reported behaviors they expected from a primary support person in response to an upcoming exam. After the exam, students and support persons independently reported behaviors the supporter provided. There was only moderate agreement within dyads about the occurrence of supportive behaviors (? = .39). Dyad intimacy, fulfilled expectancies, student and supporter social competence, and communal orientation all were associated with greater concordance. There was also greater agreement on behaviors rated by judges as helpful than harmful and in dyads that included a woman than in man–man dyads. Greater dyad intimacy and more fulfilled expectancies accounted for why there was greater concordance in dyads that included a woman, and more fulfilled expectancies accounted for why there was greater concordance on positive than negative behaviors.