This work examines the requirements for patentability in the context of biotechnology, with a special focus on the non-obviousness requirement. It analyzes non-obviousness as applied to biotechnology molecular products via a review of the relevant case law. The work begins with a typology of recombinant inventions, useful in determining ultimate non-obviousness and patentability. It distinguishes three categories of recombinant products: (1) "Translation" inventions, obtained by entering a known molecular information into a known process, (2) "Molecular modification" products, obtained by modifying prior art molecules, and (3) "Combination" inventions, obtained by combining several known functional molecular units. Recognizing the risk that many translation inventions will be considered obvious upon maturation of the underlying technology, the author examines possible alternatives for protection. The author critiques and ultimately rejects the idea of lowering the non-obviousness standard, elected by the Federal Circuit in In re Deuel. The work describes several current examples of sui generis intellectual property rights and also examines a "no action" scenario, emphasizing that the rapid changes occurring in biotechnology might ultimately make the current problem obsolete. The text also addresses broader issues such as the growing secrecy in basic science and its link to the disappearance of a clear distinction between basic and applied research. Patent law practitioners, inventors and researchers in the biotech world, and their advisors should appreciate this detailed, analysis.