Paradata are data collected during the survey data collection process. Examples of paradata are call-records data in CATI surveys (time of the call, number of call attempts, call outcome), interviewer observations, keystroke data, time stamps files, response latencies, to name just a few. These data can be used to enrich questionnaire responses or to provide information that contributes to understand and improve the survey process, and ultimately the final results. One area which might benefit from the effective use of paradata is nonresponse research. Survey researchers have increased efforts to access reliable data about non-respondents with the purpose of detecting, quantifying and adjusting for nonresponse bias. For example, call-records data are available for both espondents and non-respondents and thus are prime candidate to study non-response bias. Both call-records data and interviewer observations have potential to improve non-response adjustments, since that can be predictive of the sample persons’ probability of response to a survey or to the survey variables of interest (Kreuter and Casas-Cordero 2010). This session proposes to gather and discuss experiences about the collection and use of paradata in the perspective of non-response prevention and correction both in traditional and “new” mode surveys - online surveys, mobile phone surveys, mixed-mode surveys –, in cross sectional as well as longitudinal surveys.