The association between non-steroidal anti-inflammatory drugs (NSAIDs) use and breast cancer has remained controversial. Therefore, an overall quantitative estimate of the association needs to be studied. A systematic review and meta-analysis was executed to explore the pooled estimate for relative risk (RR) and 95% confidence interval (CI) using random or fixed effects models based on heterogeneity analysis. Overall 26 studies with 528,705 participants were included. The RR of NSAIDs use and the incidence of breast cancer was 0.94 (95% CI: 0.88-1.00) with random effects model. A slight reduction of breast cancer by taking aspirin and ibuprofen was both observed with pooled RR of 0.91 (95% CI: 0.83-0.98) and 0.81 (95% CI: 0.67-0.97), respectively. Our results indicate that NSAIDs use is associated with a slight decrease for the development of breast cancer with a marginally statistical significant difference. The associations are slightly more obvious in aspirin and ibuprofen use.