Rolling cross-section surveys (RCS) have become increasingly popular in recent years and have been implemented in many countries – also in Switzerland in the run-up to the 2011 elections.1 In an RCS, a given number of interviews are conducted every day during the entire campaign period. New addresses are added every day and the treatment of the addresses and the field work are designed in a way that interviews conducted every day constitute a random sample. The design offers many advantages for studying the influence of election campaigns on citizens’ voting behaviour, but it also provides some methodological challenges. While the theoretical concepts for using such a design are well known, little has been published on the actual fieldwork of RCS studies so far. This working paper helps to fill this gap by describing in detail the design, the fieldwork, and the methodological and practical challenges encountered during the implementation of the 2011 Swiss RCS study. We show that although there was no previous experience with the design in Switzerland, and despite some inevitable difficulties with the fieldwork, it was possible to gather high quality data. However, different challenges remain. Our experience shows some of the pitfalls to avoid and questions to be tackled when implementing such as survey. Since such information has been until now almost non-existent, we believe this might be a valuable contribution to the literature.