The Institute of Medicine makes the case that patient safety data are a critical input for redesigning care processes in ways that will make health care safer. Mandatory and voluntary error-reporting systems are sources of such data. However, a chasm of legal and practical problems exists between the collection of error reports and responding to reporting providers to improve the quality and safety of the systems in which they work. Between 2000 and 2004, the American Academy of Family Physicians (AAFP) developed and tested a voluntary error-reporting system. In this paper we discuss the current design of the AAFP's system and the legal and practical constraints that stand in the way of its becoming a more robust quality-improvement tool. We explain decisions to make the reporting system Web-based (rather than paper-based), to be anonymous (rather than confidential), to not provide direct or specific feedback to reporters, and to make it capable of receiving reports of both “sentinel” events and “intensive” reporting. This paper will clarify what is currently done with error reports and how, despite current limitations, the reporting system informs and promotes a variety of other quality initiatives of the AAFP. We also highlight how this reporting system could more robustly improve patient safety and quality in health care if legislative and other remedies are implemented to bridge the existing chasm.