The main inflammatory myopathies within the myositis group include polymyositis, dermatomyositis and inclusion-body myositis (IBM). Although potentially treatable, various practical issues have an impact on the response of these conditions to therapy. The most common reason for therapeutic failure is that the treatment targets the wrong disease, often owing to poor distinction of polymyositis from difficult-to-treat mimics such as sporadic IBM, necrotizing myopathies and inflammatory dystrophies. Evidence from uncontrolled studies suggests that polymyositis and dermatomyositis respond to treatment with prednisone at least to some degree. Empirically, adding an immunosuppressive drug might offer a 'steroid-sparing' effect or perhaps additional benefit. Intravenous immunoglobulin is proven effective as a second-line agent in patients with dermatomyositis and also seems to be effective for those with polymyositis, but offers only minimal and transient benefit to a small proportion of patients with IBM. Small, uncontrolled series suggest other agents such as rituximab or tacrolimus might offer some benefit in disease refractory to the aforementioned therapies, although IBM is resistant to most therapies. Novel agents are emerging as potential treatment options for all forms of myositis. This Review highlights common pitfalls in therapy, discusses emerging new therapies, and provides a practical therapeutic algorithm.