Abstract: Objective: To investigate the expression levels of long non-coding RNA (lncRNA) H19 in the plasma of breast cancer patients and its potential value for the diagnosis of breast cancer. Methods: The levels of lncRNA H19 were determined by quantitative reverse transcription PCR (qRT-PCR) in the cancer tissues and adjacent tissues of 24 breast cancer patients, and the plasma of 97 patients with breast cancer, 30 patients with benign breast diseases and 86 healthy controls. The correlation between plasma H19 levels and clinical pathological characteristics of breast cancer was analyzed. The levels of plasma CA153 and CEA were detected by electrochemiluminescence immunoassay (ECLI), and the receiver operating characteristic (ROC) curve and multivariable logistic regression model were used to evaluate the diagnostic value of plasma lncRNA H19 in breast cancer. Results：The expression levels of H19 in cancer tissues were significantly higher than those in adjacent tissues (Z=-2.371, P=0.018). The expression levels of plasma H19 in breast cancer patients were significantly higher than those in patients with benign breast diseases (U=411, P<0.01) and healthy controls (U=2 138, P<0.01). Moreover, the expression levels of plasma H19 in breast cancer patients correlated with the expression levels of estrogen receptor (ER), progesterone receptor (PR) and c-erbB-2, and with lymph node metastasis (all P<0.05). The area under the ROC curve (AUC), sensitivity and specificity of plasma H19 in diagnosis of breast cancer were 0.827, 57.7% and 86.4%, respectively, and its diagnostic capability was significantly higher than that of CA153 (AUC^ROC=0.661) and CEA (AUC^ROC=0.524). The combined value of plasma H19, CEA and CA153 (AUC^ROC=0.853) was higher than that of plasma H19 alone. Conclusion: Plasma lncRNA H19 may serve as a potential biomarker for the diagnosis of breast cancer.