Background: Breast cancer is the most common cancer among women worldwide. The aim of this study was to investigate the relationship between tumor size and axillary lymph node involvement (ALNI) in patients with invasive lesions, to find the best candidates for a full axillary dissection. Additionally, we evaluated the association between tumor size and invasive behavior. The study was based on data from 789 patients with histopathologically proven invasive breast cancer diagnosed in Shohada University hospital in Tehran, Iran (1993-2009). Cinical and histopathological characteristics of tumors were collected. Patients were divided into 6 groups according to primary tumor size: group I ($0.1-{leq}1cm$), II ($1.1-{leq}2cm$), III ($2.1-{leq}3cm$), IV ($3.1-{leq}4cm$), V ($4.1-{leq}5cm$) and VI (>5cm). The mean(${pm}SD$) size of primary tumor at the time of diagnosis was $3.59{pm}2.69$ cm that gradually declined during the course of study. There was a significant correlation between tumor size and ALNI (p<0.001). A significant positive correlation between primary tumor size and involvement of surrounding tissue was also found (p<0.001). The mean number of LNI in group VI was significantly higher than other groups (p<0.05). We observed more involvement of lymph nodes, blood vessels, skin and areola-nipple tissue with increase in tumor size. We found 15.3% overall incidence of ALNI in tumors ${leq}2cm$, indicating the need for more investigation to omit full axillary lymph node dissection with an acceptable risk for tumors below this diameter. While in patients with tumors ${geq}2cm$, 84.3% of them had nodal metastases, so the best management for this group would be a full ALND. Tumor size is a significant predictor of ALNM and involvement of surrounding tissue, so that an exact estimation of the size of primary tumor is necessary prior to surgery to make the best decision for management of patients with invasive breast cancer.