Purpose: To determine if the number of axillary nodes removed is a predictor of recurrence in node negative breast cancer. Materials and methods: Five hundred thirty-six patients with T1-T2, N0 invasive breast cancer, treated with lumpectomy and axillary node dissection (AND), were reviewed from January 1, 1986 to December 31, 1992. Patients received radiation to whole breast only, without regional nodal radiation. There was no adjuvant chemotherapy or Tamoxifen given. Patients were grouped according to the number of axillary nodes dissected as follows: 1-5 nodes (91 patients), 6-10 nodes (225 patients) and > 10 nodes (220 patients). Hazard ratios and p-values were determined for time to local recurrence, regional recurrence and for disease specific survival. Results: Median follow-up was 11.2 years. The overall local recurrence and regional recurrence rates for the three groups were: 1-5 nodes, 9.9% and 8.8%, respectively, 6-10 nodes, 10.2% and 2.2%, respectively, and > 10 nodes, 11.8% and 2.7%, respectively. The effect of number of axillary nodes removed was statistically significant only for regional recurrence (p = 0.017). There was no adverse effect on disease specific survival (p = 0.363). Conclusion: The number of axillary nodes removed predicts only for regional recurrence in node negative breast cancer patients, with less than 6 nodes removed associated with higher regional recurrence. This may have clinical implications with the current practice of sentinel node biopsy (SNB) replacing axillary node dissection in early stage breast cancer.