Three studies of the effects of enhanced field efforts in face-to-face surveys show that these efforts can substantially reduce the noncontact rate and result in a considerable number of converted refusers. However, evidence also seems to indicate that the ‘additional’ respondents belong to groups that were over-represented anyway. As a consequence, the final sample may be even more imbalanced than in a survey with a lower response rate, and final nonrespondents may differ substantially from late contacts and converted refusers. The paper describes these surveys, their designs and efforts to improve them. It focuses on the measures taken to enhance response rates and the results of these measures. Central are the analyses of the differences between easy and difficult respondents and possibly nonrespondents, and the question whether enhanced response rates reduce nonresponse bias.