A survey's response rate—defined as the number of completed interviews divided by the number of eligible units in the sample—is often taken to be the primary measure of quality when assessing the validity of survey data or comparing different surveys. However, the response rate suffers from several major limitations: the estimated response rate (1) is based on the sample, not the population, and so does not account for the sampling frame's undercoverage of the target population; (2) can be sensitive to the choice of e, the assumed proportion of eligible units among the units for which eligibility status has not been determined; (3) can be sensitive to the definitions used when assigning final disposition codes; and (4) does not account for misclassification of eligible units as ineligible. Because of these limitations, the response rate may be inadequate for comparing different surveys’ ability to identify and complete interviews for the target population. In recent years, this inadequacy has become more apparent due to an increase in sampling-frame options that have differing and often unknown coverage properties, and to an increase in the use of data-collection modes with differing nonresponse mechanisms. In this article, I describe a “realization” rate—a frame-independent, assumption-free measure of a survey's ability to identify and interview the target population, based on an external measure of the size of that population. I argue that the realization rate is superior to the response rate for comparing surveys that use different sampling frames or modes of data collection.