Postal questionnaires are widely used to collect data in health research and epidemiologic studies. One problem related to mail surveys is the availability of an up-to-date and accurate list of people in the population from which to draw the sample for the survey. For the identification of incorrect postal addresses it is important that all incorrectly addressed mails are returned as undeliverable. This study examines the proportion of unreturned postal letters that were sent to incorrect addresses. We sent 339 letters to existing addresses throughout Germany, but used fictional names name of persons. Three hundred and three letters (98.2%) were returned as undeliverable. The return rates only slightly differed by layout of the envelopes, region and deliverer.