In this paper, we consider the pricing of double-barrier options under a Markov-modulated regime switching diffusion model. The proposed model incorporates structural changes in economic conditions and business and investment environments into the diffusion process, which captures some important stylized facts on asset returns such as asymmetry and heavy tail. Under the proposed model, we for the first time provide the closed-form approximation to double-barrier options, which can be also further improved systemically. The numerical results and the empirical example present that our pricing solution is highly effective and our work substantially extends the existed studies.