Importance: Addition of immune checkpoint inhibitors to anti-ERBB2 treatment has shown synergistic efficacy in preclinical studies and is thus worth investigating as a neoadjuvant treatment to maximize efficacy and to minimize toxic effects. Objective: To determine if neoadjuvant atezolizumab, docetaxel, trastuzumab, and pertuzumab therapy for ERBB2-positive early breast cancer warrants continuation to the next phase. Design, Setting, and Participants: This nonrandomized, open label, multicenter, phase 2 trial was conducted by the Korean Cancer Study Group and enrolled patients across 6 institutions in Korea from May 2019 to May 2020. Eligible patients were diagnosed with ERBB2-positive breast cancer (primary tumor size >2 cm or pathologically confirmed lymph node-positive cancer, without distant metastases) with a clinical stage of II or III. Interventions: Patients received 6 cycles of neoadjuvant pertuzumab (840 mg at first cycle, 420 mg during subsequent cycles), atezolizumab (1200 mg), docetaxel (75 mg/m2), and trastuzumab (600 mg via subcutaneous injection) every 3 weeks, followed by surgery. Patients with pathologic complete response (pCR) received 12 cycles of adjuvant atezolizumab, trastuzumab, and pertuzumab every 3 weeks after surgery. Patients without pCR were treated with 14 cycles of atezolizumab, 1200 mg, plus trastuzumab emtansine, 3.6 mg/kg, every 3 weeks. Main Outcomes and Measures: The primary end point was pCR rate, which was defined as the absence of invasive cancer cells in the primary tumor and regional lymph nodes (ypT0/isN0). Secondary end points included clinical objective response rate, 3-year event-free survival rate according to pCR achievement, disease-free survival, overall survival, toxic effects, and quality-of-life outcomes. Results: A total of 67 women (median [range] age, 52 [33-74] years) were enrolled. Hormone receptor expression was positive in 32 (48%) patients. Curative surgery was performed in 65 patients because 2 patients showed disease progression during neoadjuvant treatment and their tumors became unresectable. The overall pCR rate was 61% (41 of 67 patients). The pCR rate was higher in hormone receptor-negative disease vs hormone receptor-positive disease (27 of 35 [77%] patients vs 14 of 32 [44%] patients) and in programmed cell death 1-positive expression vs programmed cell death 1-negative expression (13 of 13 [100%] patients vs 28 of 53 [53%] patients). Grade 3 and 4 neutropenia and febrile neutropenia occurred in 8 (12%) patients and 5 (8%) patients, respectively. Grade 3 and 4 immune-related adverse events occurred in only 4 patients (grade 3 skin rash, encephalitis, hepatitis, and fever). No treatment-related death occurred during the neoadjuvant phase. Conclusions and Relevance: In this nonrandomized clinical trial, treatment with the neoadjuvant atezolizumab, docetaxel, trastuzumab, and pertuzumab regimen in patients with stage II or III ERBB2-positive breast cancer appears to have had an acceptable pCR rate and modest toxic effects. Further investigation of this immunotherapy combination in ERBB2-positive early breast cancer is warranted. Trial Registration: ClinicalTrials.gov Identifier: NCT03881878.