The present study represents a replication and extension of Kivlighan et al.'s (2021) study, focusing on the social microcosm hypothesis, which posits that group members' interpersonal relationships, operationalized by intimate behaviors such as expressing anger or caring, inside the group, mirror their interpersonal relationships outside of the group. We examined the reciprocal associations between a group member's (e.g., actors) and the other group members' (e.g., partners) in-session and intersession intimate behaviors. The participants were 122 Italian graduate students (89.3% identifying as women) participating in eight-session interpersonal growth groups led by six experienced group therapists. Before each session group members completed the Interpersonal Relations Scale Checklist (Shadish, 1984) indicating their intersession intimate behaviors in the previous week. After each session group members completed the Interpersonal Relations Scale Checklist indicating their own in-session intimate behaviors. We used dynamic structural equation modeling to examine the reciprocal, temporal associations among group members' intersession and in-session intimate behaviors. The replication hypotheses were not confirmed. However, when partners engaged in higher than average intersession intimate behaviors in the previous week, then actors engaged in more in-session intimate behaviors in the current session. In addition, when actors and partners engaged in higher than average in-session intimate behaviors in the previous session, then actors engaged in significantly more intersession intimate behaviors in the current week. The results provide partial support for social microcosm theory, which predicts a reciprocal relationship for in-session and intersession intimate behaviors. As described above, there was a reciprocal relationship for partners' but not for actors' intimate behaviors. (PsycInfo Database Record (c) 2024 APA, all rights reserved).