The rules of intellectual property protection that safeguard ideas and materials produced by publicly and privately funded research have encouraged many medical and scientific innovations. Without the opportunity to obtain patents for new inventions and discoveries – and thus profit from their distribution and use – there would be no incentive to make the often massive investments required to search for and develop new drugs and other breakthroughs. Recent efforts to map the human genome triggered a debate about whether this system helps to realize the opportunities offered by modern molecular genetics. There are reasons to fear that science and its ability to deliver health benefits for everyone are being jeopardized by a widespread tendency to seek patent protection for biological materials too early. The issues at stake precede the quest for the human genome. In 1980, as new methods in molecular biology and genetics promised to reduce the lag between basic research and profitable inventions, America’s Congress passed the “Bayh-Dole Act”, which allowed patent protection for discoveries funded by US government agencies, such as the National Institutes of Health (NIH). Soon, patents were being issued on many research tools – including techniques for introducing DNA into cells, genetically altered experimental animals, and sequences of cloned pieces of DNA – that, although useful for developing health products, are not products themselves. The environment nurtured by this law (which set a precedent followed by other countries) yielded many benefits, including a dynamic biotechnology industry that produced powerful new ways to diagnose and treat disease. It also altered research in ways that inhibit the use of new findings. Many academic institutions created intellectual property offices to regulate the exchange of ideas and biological materials that were once shared freely among scientists. Some companies now make protected materials and methods available to outside researchers only under onerous terms. Recent efforts to seek patent protection for cloned genes, gene variants, portions of genes, DNA copies of messenger RNA, and genetically encoded proteins present an unusually perplexing problem. Many early patents were uncontroversial: the genes given patents were clearly related to the cause or treatment of disease, and the information covered by the patent was immediately applicable to the design of diagnostic tests or therapies. As efforts to clone and sequence human DNA expanded, patents were issued on far weaker grounds. America's Patent and Trademark Office (PTO) required only that the DNA in question serve as a “molecular probe” – a property attributable to most pieces of human DNA. Some of these newer gene patents upset the traditional balance between the exclusionary rights that patents grant to inventors and public disclosure of new, useful, and non-obvious inventions that might otherwise remain a secret. Instead, they reward the preliminary and, frankly, obvious work of determining DNA sequence – and so diminishing the prospect of financial return needed to stimulate the research that determines a genes’ function and utility. This research is made more unlikely when intellectual property protection is too broad. Some patents now grant rights to other functionally related genes solely because their sequences resemble that of the submitted gene; other patents cover gene functions that were only speculative when the patent application was submitted. This gives the patent holder title to applications discovered through the research of others. Patenting incomplete genes and gene variations also implies that multiple parties may hold title to all or part of the same gene, complicating the licensing of new technologies. The PTO is considering stricter criteria for gene patenting, putting more stress on whether the new inventions are genuinely useful. But the bar may need to be raised higher still. Under the new proposal, a patent could still be issued for a gene or a portion of a gene based on superficial and potentially misleading information that depends only on the similarity between the new gene and others previously described. These practices undermine the aims of patent policy. By over-valuing inventions and thus encouraging unduly restrictive licensing policies, they hinder, rather than promote, the use of protected information and materials. Why, for example, should potential licensees invest in developing commercial products when so-called “reach-through” provisions entitle holders of vague and uncertain patents to a share of the profits? Some remedial steps have been taken. The NIH endorsed a set of recommendations for promoting fair exchange of biological materials, and many institutions support the use of a Uniform Biological Materials Transfer Agreement to simplify this process for researchers. Similarly, recipients of patents or exclusive licenses occasionally agree to make important materials and methods, such as experimental animals and techniques for altering genes in cells, freely available to the not-for-profit sector. Parts of the private sector also sometimes recognize the virtues of public disclosure and open use of potentially patentable information. In 1998, for example, a group of pharmaceutical companies and a private philanthropy pooled resources to identify and disseminate markers of human variation in the hope that this would increase the likelihood that further work will be done to determine the significance of the markers. Achieving the right balance of protection and access to information and materials essential for basic research is essential if science is to provide practical benefits through its research. One way to move in the right direction is to provide to both the scientific community and the general public more information about the factors that influence the debate and the consequences of moving too far in either direction.