We postulate that one mechanism for the progression of breast cancers to hormone resistance involves the acquisition of mutant estrogen receptors (ER)4 by genetically unstable cell subpopulations. The T47D human breast cancer cell line may be a model for such progression, having sublines that are ER positive and estrogen responsive, ER positive and estrogen resistant, or ER negative. Also, T47D cells can be either hyperdiploid (HD) or hypertetraploid (HT) or persistently alternate between these states. T47DCO cells are a HD and ER-positive, but estrogen-resistant, subline of T47D cells that undergoes spontaneous tetraploidization. Such a stable variant, designated T47Dv, is 85% HT (Cancer Res., 49: 3943, 1989). We now show that single-cell clones derived from the mixed HD/HT T47Dv can be either HD or HT, and can be either estrogen responsive or estrogen resistant, for growth and for progesterone receptor regulation. To begin the study of ER in this model system of T47DCO and their derivatives, we have generated complementary DNA libraries from the parental HD T47DCO cells and have isolated three ER complementary DNA mutants. These include two frame-shift/termination mutants that would encode ERs truncated in the DNA-binding domain and in the hormone-binding domain and a third mutant with a large in-frame deletion spanning the hinge region and a part of the hormone-binding domain. If expressed, these mutant ERs would lack hormone-binding capacity and would be undetected by the anti-ER antibodies currently in clinical use. Genetic instability, when associated with mutant ERs in subpopulations of breast tumor cells, may provide the selective pressure leading to hormone resistance. T47DCO cells and their clonal derivatives provide a model for the systematic study of ER mutations and other mechanisms of hormone resistance in Stage IV breast cancer.