Teacher and principal surveys are among the most common data collection techniques employed in education research. Yet there is remarkably little research on survey methods in education, or about the most cost-effective way to raise response rates among teachers and principals. In an effort to explore various methods for increasing survey response rates, we randomly assigned 1,177 high school principals in the state of Michigan to 1 of 4 experimental conditions. We varied the mode of survey delivery, the mode in which the prenotification letter was sent, and whether or not a $10 incentive was provided. The results indicate that providing a monetary incentive substantially increased response rates over the no incentive condition and that principals were more likely to respond to a paper-based survey than a web-based one. (Contains 5 tables and 9 footnotes.)