Nipple-sparing mastectomy (NSM) provides a cosmetic and psychological benefit for patients, but concerns on nipple involvement (NI) of tumor continue to persist. Several studies have reported factors for predicting NI, but the results were inconsistent and uncomprehensive, making patient selection difficult. The aim of the systematic review was to pool the published data to further discern factors associated with NI. A literature review was conducted of PubMed database, following the PRISMA guidelines. Relative risks (RRs) and 95 % confidence intervals (CIs) were calculated using random-effect or fix-effect model. Publication bias and Chi-square test were also calculated. From 1978 to 2014, 27 clinical studies with 7971 patients met the inclusion criteria. Predictive factors suggest higher rates of NI including the following: tumor-to-nipple distance (TND) ≤ 2.5 cm (3.65, 1.42-9.33); positive lymph node status (2.09, 1.71-2.57); stage III or IV disease (2.41, 1.93-3.00); tumor size > 5 cm (2.42, 1.95-3.02); estrogen receptor (ER)-negative status (1.19, 1.01-1.40); progesterone receptor (PR)-negative status (1.52, 1.25-1.84); HER-positive status (1.76, 1.46-2.12); patients with ductal carcinoma in situ (DCIS) compared with invasive ductal carcinoma (1.55, 1.16-2.08). Due to the statistical heterogeneity detected with certain parameters, further investigations to confirm their association with NI will be needed. Patients with one or more risk factors such as centrally located tumors; higher tumor stage; large tumors; ER-negative/PR-negative/HER-positive status and associated DCIS have higher risk of NI. Taking these factors into consideration comprehensively may help with decision-making process for NSM.