To gain insight into the role of the CCN genes in human breast carcinomas, we quantified connective tissue growth factor (CTGF), WISP-1, CYR61, and human NOV (NOVH) mRNA expression levels in samples from 44 primary breast tumors and seven normal breasts using quantitative real-time PCR assay. Overexpression of CTGF, WISP-1, CYR61, and NOVH was found in 55 (24 of 44), 46 (20 of 44), 39 (17 of 44), and 11% (5 of 44) primary breast tumors, respectively. Statistical univariate analysis was performed to explore the links between expression of the CCN genes and clinical and pathological parameters. Interestingly, significant associations were found between CTGF expression versus stage, tumor size, lymph node status, and age at diagnosis; WISP-1 mRNA levels versus stage, tumor size, lymph node, and HER-2/neu overexpression; and CYR61 expression with stage, tumor size, lymph node, age, and estrogen receptor expression. In contrast to CTGF, WISP-1, and CYR61, no significant correlation was found between NOVH expression and any of the clinical and pathological factors. Furthermore, multivariate classification tree model analysis showed that stage and lymph node involvement were important for predicting CTGF expression in breast cancers; the stage, age, and HER-2/neu status were key factors for WISP-1 expression; and the stage, age, and estrogen receptor were valuable predictors for CYR61 expression. In summary, these results suggest that CTGF, WISP-1, and CYR61 may play a role in the progression of breast cancer and might serve as a valuable tool for monitoring tumor status of breast cancer patients.