<p>Referential process (RP) refers to how nonverbal experiences are translated into verbal forms, and it has been proposed as a foundational framework for understanding psychotherapeutic change (Bucci, 1993; Bucci, Maskit, &amp; Murphy, 2016; Bucci, 2013). While most of the prior research has examined RP in the context of psychodynamic psychotherapy and single-case designs, the current study examined RP in Cognitive Behavioral Therapy (CBT) and Emotion Focused Therapy (EFT) for 40 depressed clients. The Discourse Attributes Analysis Program (DAAP; Maskit, Bucci, &amp; Murphy, 2012) was employed to evaluate discourse attributes within 80 session transcripts. Generalized Estimating Equation (GEE) was used to analyze the data. There was a higher occurrence of affect words in EFT transcripts and a higher occurrence of reflection words in CBT transcripts reflecting each approach’s theory of change. The results of the study highlighted the therapists’ ability to effectively connect to their experiences and articulate themselves clearly to clients. Among the highest client-rated sessions, therapists of good outcome clients used more concrete, imagistic, specific, and clear language compared to therapists of poor outcome clients. Therapists of the good outcome group had higher WRADM (weighted referential activity mean) scores ((β =.018, p =.006) and WRADHP (weighted referential activity high proportion) scores (β =.059, p =.008). Moreover, clients’ session ratings increased as therapists’ WRADM (β = 15.08, p =.017) and WRADHP (β = 4.12, p =.032) increased. Contrary to the hypotheses, clients’ WRADM, WRADHP, and WRAD/WRRL did not differ between highest and lowest-rated sessions, nor did they predict clients’ session ratings. Clients in the two outcome groups also did not differ in terms of how much they symbolized and then reflected on their experiences. Clinical, theoretical, and measurement implications, limitations, and directions for future research are discussed.</p>