Objective To investigate the expression of microRNA–21 (miR–21) in breast cancer cell lines and serum of patients with breast cancer and the impact on the invasion and migration of breast cancer cells. Methods From Jan 2013 to Feb 2014, miR–21 expression were determined by fluorescent quantity polymerase chain reaction (FQ–PCR) in 4 breast cell lines (HBL–100, MCF–7, MDA–MB–231 and MDA–MB–468) and in serum from breast cancer patients (n=56), breast benign disease patients (n=39) andhealth controls (n=45). The characteristics of cell invasion and migration were examined by transwellinvasion and migration assay afterbreast cancer cell line MDA–MB–231 were transfectedwith miR–21 inhibitor or negative control by lipofectamin. The t test was used to analysis the normal distribution data. Results FQ–PCR results showed that the relative expression of miR–21 in the normal breast epithelial cell line HBL–100 was 1.01 ± 0.04, in the breast cancer cell line MCF–7, MDA–MB–231 and MDA–MB–468 were 1.99 ± 0.11, 4.02 ± 0.38 and 3.73 ± 0.79 respectively. Compared with the normal controls, miR–21 were highly expressed in the three breast cancer cell lines, the difference was statistically significant (t=9.01, 9.20 and 4.55, respectively, P<0.01); and the miR–21 was highly expressed in invasive and metastatic breast cancer cell lines (MDA–MB–231 and MDA–MB–468), compared with weakly invasive breast cancer cell line MCF–7, the difference was statistically significant (t values were 6.14 and 2.91, P<0.05), suggesting that miR–21 is highly expressed in breast cancer cells, and is closely related to the invasion and metastasis. The relative expression of miR–21 in serum of breast cancer was 2.63 (1.57–4.59), in benign breast disease group was 1.34 (1.01–1.78), in healthy control group was 0.81 (0.52–1.59), the miR–21 expression in the serum of breast cancer patients was significantly higher than in patients with benign lesions and normal control group (U values were 208 and 279, P<0.01), whereas no significant difference in serum in patients with benign lesions and normal control group, the miR–21 expression in the serum of breast cancer patients with lymph node metastasis (U=95, P=0.19) was 3.55 (2.44–5.26), significantly higher than those without lymph node metastasis [2.11(1.59–3.25), U=216, P=0.021]. The results of invasion and migration assay showed that cells treated with miR–21 inhibitor invasion was: 44 ± 18, the number of cell migration was: 98 ± 22, while the negative control treated cells after invasion was: 133 ± 44, migration cell number: 255 ± 35; miR–21 inhibitor treatment compared with the negative control, cell invasion and migration was also significantly decreased (t values were 5.46 and 9.08, P<0.01). The cell invasion and migration assay indicated the numbers of MDA–MB–231 cells, which invaded or migrated to lower chamber, were 44±18 and 98±22 respectively after miR–21 inhibitor was applied, The numbers of invaded or migrated cells were 133±44 and 255±35 when the negative control was applied. The ability of cell invasion and migration was decreased significantly in the inhibitor group compared with the negative group (tvalue separately was 5.46, 9.08, P<0.01). The capacity of breast cancer cell invasion and migrationwas significantly decreased after transfection ofmiR–21 inhibitor. Conclusions MiR–21 is highly expressed in breast cancer cell lines and breast cancer patients′ serum. Altered expression of miR–21 maybeplays an important role in breast cancer invasion and migration. MiR–21 may serve as new biomarker to early detectionand prognosis estimation of breast cancer. (Chin J Lab Med, 2015, 38: 186–190) Key words: Breast neoplasms; microRNAs; Neoplasm metastasis; Tumor markers, biological