<p>A growing body of research has found contradictions between self-report symptom inventories and other information sources when evaluating psychotherapy outcomes - a phenomenon described as a paradoxical outcome. Previous studies have highlighted the need to identify patterns of this phenomenon. This article addresses this by conducting a systematic review and meta-synthesis of 12 case studies. Using Thematic Synthesis, two analyses were performed: one identifying manifestations of paradoxical outcome, and one examining researchers’ interpretations. Three meta-themes described the phenomenon: (1) false improvement, (2) false worsening, and (3) invisible change. Across studies, variations in the therapeutic alliance were consistently linked to paradoxical outcome patterns. A fourth meta-theme − (4) the inclusion of uncaptured aspects of psychotherapeutic outcome - highlighted researchers’ interpretations, suggesting that paradoxical outcomes often reflect aspects of therapeutic change not captured by standard symptom inventories. The findings emphasize the need to move beyond linear views of symptom trajectories, to triangulate multiple sources of information, and to interpret symptom changes within the context of patients’ lived situations. The study is registered in PROSPERO (CRD42024535341).</p>