Purpose The clinical significance of lymph node micrometastases and isolated tumor cells (ITCs) in breast cancer is still controversial. After a median follow-up of 52 months, a report from the Swedish Multicenter Cohort Study presented a worse cancer-specific and event-free survival for patients with micrometastases than node-negative individuals, but could not demonstrate a significant difference in overall survival (OS). Due to the tendency of breast cancer to relapse after more than 5–10 years, we now report the long-term survival of the cohort. Methods Between September 2000 and January 2004, 3355 breast cancer patients were included in a prospective cohort. Sentinel lymph node biopsy was always performed. Patients were classified in four groups according to their overall nodal stage: node negative (N0, 2372), ITCs (113), micrometastases (123), and macrometastases (747). Kaplan–Meier survival estimates and Cox proportional hazard regression models were applied. Results Median follow-up was 156 months. Ten-year cancer-specific survival and OS were significantly lower in case of micrometastases than in N0 (84.7 vs. 93.5%, p  = 0.001, and 75.5 vs. 84.2%, p  = 0.046, respectively). In case of macrometastases, corresponding survival rates were 82.8 and 74.3%. Only for those aged less than 50 years, cancer-specific survival and OS were significantly worse in case of ITCs than N0. Patients with micrometastases received less often chemotherapy than those with macrometastases (24.4 vs. 53.9%). Conclusions Lymph node micrometastases in breast cancer have a prognostic significance. This study demonstrates a similar survival for patients with micrometastases and those with macrometastases, possibly due to systemic undertreatment.