Today, many Internet users take part in online social networks (OSNs) like Facebook, Xing, LinkedIn, MySpace, studiVZ, and others. However, the motives for participation are multifaceted and stand in contrast to the motivation not to join OSNs because of the potential danger with respect to privacy aspects. This risk is already perceived by some OSN users but still underestimated. The naive handling with personal data inside such networks combined with many possibilities to link several OSN profiles of a single user paves the way for third-parties to gather the comprehensive virtual appearance of a person. Therefore, to turn the tables, such correlations of information might be useful to demonstrate to a user a measure of how linkable his personal data is, given the current exposure of his public data in OSN profiles. In this paper we point out some astonishing simple ways to automatically link data from different OSNs. As we want to and have to act compliant to the German data protection laws, we analyze corresponding requirements and sketch our concept for compliant statistical sampling. We further show preliminary results regardi ... mehrng successful profile correlations based on, even under legal and technical constraints, extracted friends lists.