I conducted a six-month research on aerodynamic performance-driven design for skyscrapers, sponsored by RTKL’s Kagan Fellowship. What is shown here is my presentation at Computer Aided Architectural Design (CAAD) Futures 2013 conference in Shanghai. Strategies in modifying the shape of the skyscrapers with the goal of improving their aerodynamic performance have been suggested by various scientific researches. Six of these strategies are incorporated into parametric design tools that produce a matrix of 60 prototypes, which then undergo qualitative and quantitatively evaluation iterations to select the most optimized design, with considerations given to aerodynamic performance, structural robustness, program potential, and image attractiveness. Consulting from the world-renowned engineering firm RWDI and the use of Autodesk Simulation CFD are used to make the evaluation. Eventually the 60 prototypes are narrowed down to two design options, which in turn are further developed into hypothetical projects.   The most important result of the study is not the design, but a multi-staged aerodynamic performance-driven design process. In addition, two valuable insights have been obtained: first, to inspire a new approach in the design of skyscrapers, the Multi-Disciplinary Optimization methodology from the aerospace industry has been used. Second, parametric design that is able to generate interesting forms is backed by quantitative scientific analysis and reasoning. Instead of the Modernist “form follows function”, the philosophy of the new approach is “form follows performance”. This is the train of thought that I have found fascinating: aerodynamic performance, energy performance, and performances of other kinds, will inspire new design approaches. When physics is brought into architecture, the Nature's process is re-createdin a way . Look at what fantastic “architecture” the Nature is able to produce by being performance-driven (you can argue it is adaption-driven, but I would argue that natural selection should not be overemphasized in the context of the discussion.)