We have compared the cell and tissue selective estrogenic and antiestrogenic activities of tamoxifen, raloxifene, ICI 164,384 and a permanently ionized derivative of tamoxifen--tamoxifen methiodide (TMI). This non-steroidal antiestrogen has limited ability to cross the blood brain barrier and is therefore less likely to cause the central nervous system disturbances caused by tamoxifen. We have used the stimulation of the specific activity of the "estrogen induced protein", creatine kinase BB, as a response marker in bone, cartilage, uterine and adipose cells and in rat skeletal tissues, uterus and mesometrial adipose tissue. In vitro, TMI, tamoxifen and raloxifene mimicked the agonistic action of 17beta-estradiol in ROS 17/2.8 rat osteogenic osteosarcoma, female calvaria, and SaOS2 human osteoblast cells. In Ishikawa endometrial cancer cells, tamoxifen showed reduced agonistic effects and raloxifene showed no stimulation. However, as antagonists, tamoxifen and raloxifene were equally effective in Ishikawa or SaOS2 cells. In immature rats, all four of the antiestrogens inhibited estrogen action in diaphysis, epiphysis, uterus and mesometrial adipose tissue; when administered alone, tamoxifen stimulated creatine kinase (CK) specific activity in all these tissues. Raloxifene and TMI, however, stimulated only the skeletal tissues and had no stimulatory effect in the uterus or mesometrial fat, and the pure antiestrogen ICI 164,384 showed no stimulatory effect in any of the tissues. The simultaneous injection of estrogen, plus an antiestrogen which acted as an agonist, resulted in lower CK activity than after injection of either agent alone. These differential effects, in vivo and in vitro, may point the way to a wider therapeutic choice of an appropriate antiestrogen which, although antagonizing E2 action in mammary cancer, can still protect against osteoporosis and cardiovascular disease and not stimulate the uterus with its attendant undesirable changes, or interfere with the beneficial action of E2 in the brain.