Preceding the recruitment for a new, online household panel a pilot study was carried out designed to find the optimal recruitment strategy. A probability sample of addresses was drawn for this study, including households without Internet access and households without a fixed landline. Households without Internet access are provided with Internet and – if necessary – with a computer by the survey institute to be able to participate in the online panel. All households were contacted in a traditional way: first, an announcement letter was sent, explaining the nature of the study; next, respondents were contacted by an interviewer, either by telephone or face-to-face. The interviewer asked them to participate in a 10-minute interview, after which, at the end of the interview, the request to participate in the panel was made. The 10-minute recruitment interview was designed, on the one hand, to give respondents an impression of what the panel interviews would be like, and, on the other hand, to measure the characteristics of respondents who agree and disagree to participate in the panel. Of special interest were characteristics other than the standard demographics, which can often be corrected by weighing, and therefore the interview included indicators of health; social integration; political interest; leisure activities; survey attitudes; loneliness and personality. Results show, firstly, important effects of age and prior access to Internet. These related effects indicate that for the real recruitment, considerable extra effort has to be put into persuasion of the elderly who are reluctant towards the free Internet access and computer offered by the survey institute. Furthermore, the respondents who were willing to become a panel member were, on average, more interested in the news; made more cinema or museum visits; were more sportive or participating in unions or clubs; more active voters; more self-confident; less lonely; less religious; and more positive towards survey research in general than the respondents who were unwilling to become a panel member. The rapidly growing subpopulation of respondents without fixed landline can be characterised as of younger age; single person households; working; much less than average involved in news, politics and societal activities; often non-voters; somewhat more often speaking another language than Dutch at home.