Quantitative survey research requires micro data. If the target population is large, sampling is necessarily the only option. There exists a general acceptance that sampling should be probability based. The available sampling frame may however be inadequate and non-response may be high and selective. These nui- sances are becoming more common for many reasons. An important reason is the survey mode that is more often a telephone or a web since these are cheaper and more comfortable. However, the frames for such surveys are more restricted than in the case of face-to-face interviewing. Non-response has been generally increasing over recent decades. Moreover, multinational surveys have been becoming more common, and the success in sampling and data collection will have a crucial role in order to get comparable estimates. Also, longitudinal approaches are often desired to follow individuals over time. The session has invited contributors who share these challenges and who have new experiences in these areas. We have received a good number of papers and several above questions are considered in these papers. The introduction rst gives an overview about challenges and nuisances in today and future survey research from the point of view of sampling, and then briey summarises the papers presented in the following three sessions.