Spinal and bulbar muscular atrophy (SBMA) is a late-onset, progressive neurodegenerative disease linked to a polyglutamine (polyQ) expansion in the androgen receptor (AR). Men affected by SBMA show marked muscle weakness and atrophy, typically emerging midlife. Given the androgen-dependent nature of this disease, one might expect AR antagonists to have therapeutic value for treating SBMA. However, current work from animal models suggests otherwise, raising questions about whether polyQ-expanded AR exerts androgen-dependent toxicity through mechanisms distinct from normal AR function. In this study, we asked whether the nonsteroidal AR antagonist flutamide, delivered via a time-release pellet, could reverse or prevent androgen-dependent AR toxicity in three different mouse models of SBMA: the AR97Q transgenic (Tg) model, a knock-in (KI) model, and a myogenic Tg model. We find that flutamide protects mice from androgen-dependent AR toxicity in all three SBMA models, preventing or reversing motor dysfunction in the Tg models and significantly extending the life span in KI males. Given that flutamide effectively protects against androgen-dependent disease in three different mouse models of SBMA, our data are proof of principle that AR antagonists have therapeutic potential for treating SBMA in humans and support the notion that toxicity caused by polyQ-expanded AR uses at least some of the same mechanisms as normal AR before diverging to produce disease and muscle atrophy.