This study examined the within-dyad association of immediacy (i.e., a skill that therapists use to work with the therapeutic relationship in the here-and-now) with the strength and congruence of the working alliance across 1352 sessions of 58 adult community clients seeing 11 doctoral student therapists in individual psychodynamic-interpersonal psychotherapy. As a preliminary step, the factor structure and validity were tested for the Metacommunication in Session Questionnaire–Client Form (MSQ-C), a client-rated measure of immediacy adapted from the supervisory MSQ (Calvert, Deane, & Grenyer, 2020). After every session, clients and therapists completed the Working Alliance Inventory–Short Revised (WAI-SR; Hatcher & Gillaspy, 2006) and clients completed the MSQ-C. Factor analysis supported a two-factor structure for MSQ-C (Open Communication and Managing Disagreement/Discomfort factors). Validity of the MSQ-C was supported by predicted correlations with measures of helping skills, sessions quality, alliance, and therapist reported immediacy use, although some associations varied depending on the client or therapist rater perspective. Results of multilevel, latent variable models found that when clients reported more immediacy in a session compared to their average session, they tended to report a stronger alliance; and this effect was strongest in earlier sessions, weaker in magnitude in middle sessions, and non-significant in later sessions. Results of multilevel truth-and-bias models showed that therapist alliance ratings were temporally congruent with client alliance ratings, but client-perceived immediacy did not predict alliance congruence. Limitations and future directions are discussed.