The status of the 66-kDa human estrogen receptor-alpha (hER-alpha66) is a critical determinant in the assessment of the prognosis and in the design of treatment strategies of human breast cancer. Recently, we cloned the cDNA of an alternatively spliced variant of hER-alpha66, termed hER-alpha36; the predicted protein lacks both transcriptional activation domains of hER-alpha66 but retains its DNA-binding domain, partial dimerization, and ligand-binding domains and three potential myristoylation sites located near the N terminus. These findings thus predict that hER-alpha36 functions very differently from hER-alpha66 in response to estrogen signaling. We now demonstrate that hER-alpha36 inhibits the estrogen-dependent and estrogen-independent transactivation activities of hER-alpha66 and hER-beta. We further demonstrate that hER-alpha36 is predominantly associated with the plasma membrane where it transduces both estrogen- and antiestrogen-dependent activation of the mitogen-activated protein kinase/extracellular signal-regulated kinase signaling pathway and stimulates cell growth. We conclude that hER-alpha36 is a predominantly membrane-based, unique alternatively spliced variant of hER-alpha66 that acts as a dominant-negative effector of both estrogen-dependent and estrogen-independent transactivation functions signaled through hER-alpha66 and ER-beta; it also transduces membrane-initiated estrogen-dependent activation of the mitogen-activated protein kinase/extracellular signal-regulated kinase mitogenic signaling pathway. The estrogen and antiestrogen signaling pathways mediated by hER-alpha36 provide an alternative explanation for why some human breast cancers are resistant to and others are worsened by antiestrogen therapy; the data suggest that hER-alpha36 also may be an important marker to direct therapy in human breast cancers, and perhaps hER-alpha36 also may transduce estrogen-dependent signaling in other estrogen target tissues.