BACKGROUND Increased soluble human epidermal growth factor receptor 2 (sHER2) is an indicator of a poor prognosis in HER2-positive metastatic breast cancer. In this study, the authors evaluated levels of sHER2 during treatment and at the time of disease recurrence in the adjuvant North Central Cancer Treatment Group N9831 clinical trial. METHODS The objectives were to describe sHER2 levels during treatment and at the time of recurrence in patients who were randomized to treatment arms A (standard chemotherapy), B (standard chemotherapy with sequential trastuzumab), and C (standard chemotherapy with concurrent trastuzumab). Baseline samples were available from 2318 patients, serial samples were available from 105 patients, and recurrence samples were available from 124 patients. The cutoff sHER2 value for the assay was 15 ng/mL. Statistical methods included repeated measures linear models, Wilcoxon rank-sum tests, and Cox regression models. RESULTS There were differences between groups in terms of age, menopausal status, and hormone receptor status. Within treatment arms A, B, and C, patients who had baseline sHER2 levels ≥15 ng/mL had worse disease-free survival than patients who had baseline sHER2 levels <15 ng/mL (arm A: hazard ratio, 1.81; P = .0014; arm B: hazard ratio, 2.08; P = .0015; arm C: hazard ratio, 1.96; P = .01). Among the 124 patients who experienced disease recurrence, sHER2 levels increased from baseline to the time of recurrence in arms A and B but remained unchanged in arm C. Patients who had recurrence sHER2 levels ≥15 ng/mL had a shorter survival after recurrence with a 3-year overall survival rate of 51% compared with 77% for those who had recurrence sHER2 levels <15 ng/mL (hazard ratio, 2.36; 95% confidence interval, 1.19-4.70; P = .01). CONCLUSIONS In patients with early stage, HER2-positive breast cancer, a high baseline sHER2 level was identified as a prognostic marker associated with shorter disease-free survival, and a high sHER2 level at recurrence was predictive of shorter survival.