Driven by the increasing bike-sharing programs, demand for urban cycling has increased rapidly in recent years. As one of the effective ways to encourage cycling, bikeway has been recognized to support the cycling commutes. Therefore, the rational planning of the bikeway routes and the construction of bikeway network is crucial for improving cycling commutes in urban areas. Taking Beijing as the case, this study proposes an analytical framework for selecting and optimizing bikeway routes based on mobile phone data. The cycling demand is analyzed by mining the mobile phone geographic information, followed by evaluating the available road spaces for bikeway by applying the TOPSIS approach. Finally, the study applies the least-cost path method to select the bikeway routes, which are then classified by cycling potentials and traffic volumes for two scenairos of bikeway construction, respectively. The results show that (1) In city center, the bikeway are arranged to serve for cycling commutes within the central urban area, while the bike lanes in the suburbs (e.g., Haidian, Chaoyang, and Yizhuang) primarily serve for inter-district commute travels; (2) Bikeways that meet for the demand of interdistrict commutes should be given priority to construct and is suggested to be as bicycle expressways; (3) A total length of 450 km bikeways is planned to form a city-wide network, which covers more than 70% commute population and primarily connects the high-density suburbs with central areas. The results are expected to provide references for urban planners and local governments to reasonably plan the bikeway network.