Traditional modes of data collection like face-to-face and telephone are facing strong challenges (response rates going down, costs going up). At the same time, the web provides attractive features for doing surveys: it can be very quick, can reach people all over the world, and may be done in a quite easy way. Therefore, in the past 10 years, web surveys became more and more numerous. However, the representativeness and the quality of the data obtained via these web surveys may be questioned. Because it is easy to make a web survey, lots of people started doing their own surveys without taking the time to think about methodological issues. The issues of web coverage of the population studied and of self-selection and non-response are very important ones. The issue of measurement errors is also crucial. The panel will focus mainly on this last point: assessment of the quality of web surveys. language of the panel: English (but could be changed to Spanish if the presenters prefer).