The development of chemoresistant breast cancer is poorly understood and second treatment options are barely investigated. The term 'chemoresistance' is ill-defined and thus, our experimental analyses aimed to disentangle the resistance to cell cycle arrest from the resistance to trigger apoptosis, both of which are important mechanisms to be targeted by anticancer therapy. Therefore, an MCF-7 array, which encompassed clones harboring distinct genetically- and pharmacologically-induced stages of resistance, was established. For this, MCF-7 cells were stably transfected with erbB2 cDNA and a dominant negative p53 mutation and the two clones were subjected to long-term treatment with the clinical agents 2'-deoxy-5-fluorouridine (5-FdUrd) or arabinosylcytosine (AraC) to develop specific chemoresistance. This array was tested with 3,4',5-trihydroxy-trans-stilbene (resveratrol) and the methoxylated paired stilbene analogue 3,4',5-trimethoxy-trans-stilbene (M5) to investigate whether these agents can overcome genetically- and pharmacologically-induced chemoresistance and to correlate the structure-activity relationship of resveratrol and M5. In all conditions tested, M5 exhibited stronger anticancer activity than resveratrol, but the cell cycle inhibitory properties of the tested drugs were dependent on the genetic background and the chemoresistant phenotype. In contrast, the proapoptotic properties were rather similar in the distinct genetic backgrounds of the clone array and therefore, apoptotic triggers and cell cycle checkpoints were distinctly affected and are thus independent of each other. The study demonstrates the merits or virtues of the genotypically- and phenotypically-defined clones of the MCF-7 array as a testing tool for novel drugs, which discriminates the two types of chemoresistance mechanisms.