Vertically integrated systems of health care share a set of interests with academic health services researchers. These interests include questions about: the appropriate size and scope of the various system components; mechanisms for coordinating care across component parts; and "what works" in terms of clinical effectiveness within components. Primary care is a key system component in which all of these questions apply. Traditional research designs, however, are often difficult to apply in a health system context. The problems are (1) the rapid pace at which organizational change occurs, which renders some research findings obsolete before they are published, and (2) the difficulty in generalizing across a complex set of local environmental factors that vary from system to system. Useful research from a system perspective will either focus on problems of wide generalizability over place and time or be conducted according to new methodologies that offer much faster "cycle time" from design to findings.