This paper outlines ways to maximize response rates to surveys by summarizing the most relevant literature to date and demonstrating how these techniques have resulted in consistently high rates of return in family practice research. We describe the methodology used in recent surveys of physicians conducted by the Centre for Studies in Family Medicine through its Thames Valley Family Practice Research Unit, located in London, Ontario, Canada and funded by the Ontario Ministry of Health and Long-Term Care. The identification and implementation of these techniques to maximize response rates is critical, as primary health care researchers often rely on information gathered through questionnaires to study physicians' practice profiles, experiences and attitudes. Four separate and distinct mailed surveys of physicians using a modified Dillman approach were conducted from 2001 to 2004. The sampling strategies, topics, types of questions and response formats of these surveys varied. The first survey did not use any incentives or recorded delivery/registered mail and received a response rate of 48%. In sharp contrast, the other three surveys obtained responses rates of 76%, 74%, 74%, respectively, achieved through the use of gift certificates and recorded delivery/registered mail. Sending a survey by recorded delivery/registered mail tends to result in the survey package being given priority in the physicians' incoming mail at the practice. Gift certificates partially compensate physicians for time spent completing the survey and recognition of the time required is appreciated. The response rates achieved provide strong evidence to support the use of monetary incentives and recorded delivery/registered mail (along with the Dillman approach) in survey research. It is anticipated that this evidence will be used by other researchers to justify requests for funding to cover the costs associated with incentives and recorded delivery/registered mail. We recommend the use of these strategies to maximize response rates and improve the quality of this type of primary health care research.