In most of countries, patent law system provides the patent holder with exclusive right upon his/her invention; anyone who uses the patented invention without patentee's consent shall be liable for patent infringement. However, the system also establishes certain exceptions (or defenses) due to the public policy, such as experimental use, prior users…etc. Under the American patent system, experimental use defense is a common law doctrine, it was originated in Justice Story's 1813 opinion in Whittemore v. Cutter. Since then, it has been successfully asserted as a legitimate defense at numerous patent infringement cases. 1995, the Congress established experimental exemption for generic drug industries under 35 U.S.C. §271(e)(l). In the academic society, utilizing patented invention to do the research by professors/researchers has never been questioned until Madey brought patent infringement lawsuit against Duke University in 1998. And, in 2002, Federal Circuit rejected an experimental use defense asserted by Duke University. It is still unknown how the decision may affect academic research as well as industrial society. However, it shall be noted that Bayh-Dole Act of 1980 which encourages the establishment of university-industry research relationship causes most of university researches, even basic ones, lead to commercial interests. And, in the field of biotechnology, it is inevitable for researchers (either from university or industry) to take experiments on patented inventions so as to realize their technique and functions. According to Madey case, those uses constitute patent infringement and the experimental exemption would not be applied. Would that be the case, does it help the industrial development due to the thorough protection on patent rights, or does it harm the industrial development due to discouragement on scientific research. Some writers suggested broad experimental exemption so as to promote industrial technology. Others were concerned the upstream patents or patented research tools may diminish downstream inventions due to the strict experimental exemption. The author of this article suggests the balancing test between patent holders' right and the public interest of higher industrial technology. If the upstream patent may prevent downstream inventions, the problem should not be whether the experimental exemption is broad enough, but whether the patent is too broad that should not be issued, or the whole patent system should be reformed. Although our patent law also provides experimental use defense, it is seldom bought out at patent infringement cases, whether by universities or by industries. Owing to the growing relationship between universities and industrial society, and development of biotechnology, it's time for us to realize the definition of experimental exemption. Clause 1, Section 1 of Article 57 under our patent law does not clarify the definition of non-profit activity of research/experiment. The 2006 revised draft does establish the exemption, which consists of the conduct as provided under 35 U.S.C. § 271(e)(l) and any other experiments.