Randomized controlled trials may lack generalizability and cannot always provide sufficient information for policymakers to make decisions regarding the use of new technologies entering into a health care system. Information related to other aspects of health technologies, in addition to clinical endpoints, such as economic, patient-related, and organizational issues, may be needed before making policy decisions. To address generalizability issues related to a health care technology, primary data collection conducted within a health care system may be required. A process for primary data collection to inform health policy in Ontario regarding health care technologies was initiated in 2003. Following recommendations from the Ontario Health Technology Advisory Committee, conditionally funded field evaluations for promising technologies with insufficient information are initiated. The conduct of these field evaluations occurs in 3 stages. The translation of the policy problem to research questions that can be answered through the design of a practical clinical trial is the first stage, followed by implementation, data collection, quality control, and quality assurance activities. After the completion of the field evaluation data collection and analysis, the results of the study are presented to the committee to respond to its request. Technologies evaluated by this process include drug-eluting stents, the endovascular repair of abdominal aortic aneurysms, and multifaceted interventions in diabetes care. This process, developed in Ontario for the evaluation of health technologies, has been a successful collaborative effort between practitioners, academia, government, health policymakers, and industry that has permitted an evidence-based framework for the approval and funding of technologies within the health care system.