In patent law, an invention patent application is published early and then will have substantive examination while utility model and design patent application are only examined from format perspective. A patent application needs not only to be novel and useful but also have inventiveness. Inventiveness lies in the heart of the patent system. Inventiveness sets a more strict standa rd f or g ra nting a pa tent a nd r equires tha t a n invention as a whole should be non-obvious in light of the prior art. The original purpose of the non-obviousness makes sense in preventing those obvious inventions from enjoying monopoly and having the real invention get protected thereby promoting inventors to make their invention public to promote the development of the technology. Through the empirical analysis of patent administrative cases in our country, we find that assessing inventive step is the most prominent focus and difficulty problem, but the systematic theory and practice research on this subject matter is little. This dissertation focuses on the methods of assessing inventive step, by comparison with the United States and European patent system to find solutions. This dissertation includes three parts, the first part referred to as the theoretical part focused on the basic theoretical issues about assessing inventive step, defines inventive step and discusses its system meaning, introduces the inventive step of U.S Patent law, the European patent convention and illustrates the importance of the establishment of the system of the inventive step and significant role of inventive step criteria in patent protection. The second part discuss the mainly methods of assessing inventive step, including main influential factors and the secondary considerations. The third part presents a number of recommendations aimed to promote the objectivity, stability and unity in assessment of inventive step.