The effects of estrogen may be exerted via the nuclear estrogen receptors (ERs) ERalpha or ERbeta or via the recently proposed transmembrane estrogen receptor G protein-coupled receptor 30 (GPR-30). The purpose of this study was to elucidate the ER specificity for the ameliorating effects of estrogen on arthritis and bone loss in a model of postmenopausal rheumatoid arthritis (RA). Female DBA/1 mice underwent ovariectomy or sham operation, and type II collagen-induced arthritis was induced. Mice were treated subcutaneously 5 days/week with the specific agonists propylpyrazoletriol (PPT; for ERalpha), diarylpropionitrile (DPN; for ERbeta), G1 (for GPR-30), or with a physiologic dose of estradiol. Clinical arthritis scores were determined continuously. At termination of the study, bone mineral density (BMD) was analyzed, paws were collected for histologic assessment, serum was analyzed for cytokines and markers of bone and cartilage turnover, and bone marrow was subjected to fluorescence-activated cell sorting. Treatment with PPT as well as estradiol dramatically decreased the frequency and severity of arthritis. Furthermore, estradiol and PPT treatment resulted in preservation of bone and cartilage, as demonstrated by increased BMD and decreased serum levels of bone resorption markers and cartilage degradation markers, whereas no effect was seen after DPN or G1 treatment. In a well-established model of postmenopausal RA, ERalpha, but not ERbeta or GPR-30 signaling, was shown to ameliorate the disease and the associated development of osteoporosis. Since long-term treatment with estrogen has been associated with significant side effects, increased knowledge about the mechanisms behind the beneficial effects of estrogen is useful in the search for novel treatments of postmenopausal RA.