Blended therapy (BT) combines digital with face-to-face psychological interventions. BT may improve access to treatment, therapy uptake, and adherence. However, research is scarce on the structure of BT models. We synthesized the literature to describe BT models used for the treatment of psychological disorders in adults. We investigated whether BT structure, content, and ratio affected treatment efficacy, uptake, and adherence. We also conducted meta-analyses to examine treatment efficacy in intervention-control dyads and associations between treatment outcomes versus BT model structure. PsycINFO, CINAHL, Embase, ProQuest, and MEDLINE databases were searched. Eligibility criteria included articles published in English till March 2023 that described digital and face-to-face elements as part of an intervention plan for treating psychological disorders in adult patients. We developed a coding framework to characterize the BT interventions. A meta-analysis was conducted to calculate effect size (ES; Cohen d and 95% CIs) regarding pre- and posttreatment outcomes in depression and anxiety versus BT structure. The review was registered with PROSPERO and followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Searches identified 8436 articles, and data were extracted from 29 studies. BT interventions were analyzed and classified according to mode of interaction between digital and face-to-face components (integrated vs sequential), role of the components (core vs supplementary), component delivery (alternate vs case-by-case), and digital materials assignment mode (standardized vs personalized). Most BT interventions (n=24) used a cognitive behavioral therapy approach for anxiety or depression treatment. Mean rates of uptake (91%) and adherence (81%) were reported across individual studies. BT interventions were more effective or noninferior to treatment as usual, with large spread in the data and a moderate to large ES in the treatment of depression (n=9; Cohen d=-1.1, 95% CI -0.6 to -1.6, P<.001, and z score=-4.3). A small, nonsignificant ES was found for anxiety outcomes (n=5; Cohen d=-0.1, 95% CI -0.3 to 0.05, P=.17, and z score=-1.4). Higher ESs were found in blended interventions with supplementary design (depression: n=11, Cohen d=-0.75, 95% CI -0.56 to -0.95; anxiety: n=8, Cohen d=-0.9, 95% CI -0.6 to -1.2); fewer (≤6) face-to-face sessions (depression: n=9, Cohen d=-0.7, 95% CI -0.5 to -0.9; anxiety: n=7, Cohen d=-0.8, 95% CI -0.3 to -1.3); and a lower ratio (≤50%) of face-to-face versus digital sessions (depression: n=5, Cohen d=-0.8, 95% CI -0.6 to -1.1; anxiety: n=4, Cohen d=-0.8, 95% CI 0.006 to -1.6). This study confirmed integrated BT models as feasible to deliver. We found BT to be effective in depression treatment, but anxiety treatment results were nonsignificant. Future studies assessing outcomes across different psychological disorders and therapeutic approaches are required. PROSPERO CRD42021258977; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=258977.