Three-dimensional quantitative structure-activity relationships (3D-QSARs) on the inhibitory activity of [(2-phenylindol-3-yl)methylene]propanedinitrile analogues ($1{sim}19$) against human breast cancer cells (MDA-MB 231) were studied using comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) methods. The optimized CoMFA model 2 (${r^2}_{cv.}(q^2)$=0.581, ${r^2}_{ncv.}$=0.970) model predictability was lower than that of CoMSIA model 2, but showed better fitness than the CoMSIA model 2 (${r^2}_{cv.}(q^2)$=0.970, ${r^2}_{ncv.}$=0.886). The contour maps showed that, the inhibitory activities of the analogues against breast cancer cells were expected to increase when hydrophilic and steric favor groups with less than five carbon atoms were substituted at the $R_1$ position. However, it was predicted that the negative charge $R_2$ favor group and hydrophobic favor, along with the positive charge favor and steric $R_3$ disfavor group will achieve the inhibitory activity. The inhibitory activity ($IC_{50}$=0.0018 ppm) against breast cancer cells of the newly designed molecule (P1) with optimized CoMFA model 2 was 20-fold higher than that of the commercialized drug, Docetaxel ($IC_{50}$=0.04 ppm).