The aim of this work was to predict the anticancer potential of 3 atisane, and 3 trachylobane diterpene compounds extracted from the roots of Xylopia langsdorffiana. The prediction of anticancer activity as expressed against PC-3 tumor cells was made using a PLS model built with 26 diterpenes in the training set. Significant statistical measures were obtained. The six investigated diterpenes were applied to the model and their activities against PC-3 cells were calculated. All the diterpenes were active, with atisane diterpenes showing the higher pICso values. In human prostate carcinoma PC-3 cells, the apoptosis mechanism is related to an inhibition of IKK/NF-KB. Antioxidant potential implies a greater electronic molecular atmosphere (increased donor electron capacity), which can reduce radical reactivity, and facilitate post donation charge accommodation. Molecular surfaces indicated a much greater electronic cloud over atisane diterpenes.