ABSTRACT In Pakistan, breast cancer (BC) is frequently diagnosed with advanced disease. We aimed to examine the association of breast tumor size with reasons of reporting delays, care intervals in patients with advanced disease. This cross-sectional study enrolled 392 BC patients from tertiary care hospitals. Data were collected from August 2018–March 2019. Chi-square for significance and logistic regression for association were used. Patients between 31 and 45 years of age (51%), rural residents (39%), lower-class (88.4%), no family history (17.1%), at stage-IV (36%) and patient interval >90 days (70%, p = .034) presented with large tumor size (>5 cm). Tumor size was significantly associated with area of residence (p = .043), social-class (p = .027), family history (p = .004), smoking (p = .021), nipple discharge (p = .004), recurrence (p = .024), and metastasis (p = .007). Patient-interval was associated with poor knowledge (OR;4.3,p = .0001), influence of traditional healers (OR;2.3,p = .05), religion (OR;3.9,p = .0001), finances (OR;2.4,p = .045), and competing life priorities (OR;2.9,p = .026). In-adjusted linear regression model, area of residences, education, social-class, family-history, recurrence, cancer type, and patient interval (β;0.110,p = .030) were found to be independent predictors of tumor size. In conclusion, education, family history, area of residence, social class, recurrence, cancer stage and patient interval, co-influenced by religion, finances, life priorities, traditional healers, and poor knowledge, were independent predictors of tumor size in BC patients.