Background: The ability to predict the behaviour of breast cancer from its dimensions allows the clinician to inform a woman about the absolute benefits of adjuvant therapies or further surgery to control her disease. Tumour size and grade are independent predictors of nodal disease. This study aims to generate a tool, using Australian data, allowing surgeons to calculate the probability of axillary lymph node involvement in a preoperative setting. Methods: The histological reports of patients with breast cancer treated in 1995 in New South Wales were examined and tumour size, grade and nodal status recorded. Univariate and multivariate analyses identified predictors of node positivity and, using linear regression analysis, a simple formula to predict nodal involvement was derived. Results: In a 6-month period, 754 women had non-metastatic, unifocal breast cancer treated with surgery and complete axillary dissection and 283 (37.5%) had positive nodes. Tumour size remained an independent predictor of node positivity and the probability (%), y, of nodal involvement may be predicted by the formula y = 1.5 × tumour size (mm) + 7, r = 0.939 and P = 0.001. Conclusion: This paper shows the need to assess the axilla in every patient because even patients with small tumours (0-5 mm) have the possibility of axillary involvement (7-14.5%). Use of this simple formula allows clinicians and patients to make informed decisions about the possible need for a full axillary dissection to reduce the chance of understaging and potentially undertreating a woman's breast cancer.