In asylum decision-making, legal authorities rely on the criterion "credibility"as a measure for determining whether an individual has a legitimate asylum claim; that is, whether they have a well-founded fear of persecution upon returning to their country of origin. Nation states, international institutions, and NGOs increasingly seek to leverage data-driven technologies to support such decisions, deploying processes of data cleaning, contestation, and interpretation. We qualitatively analyzed 50 asylum cases to understand how the asylum decision-making process in Denmark leverages data to configure individuals as credible (or not). In this context, data can vary from the applicant's testimony to data acquired on the applicant from registers and alphanumerical data. Our findings suggest that legal authorities assess credibility through a largely discretionary practice, establishing certainty by ruling out divergence or contradiction between the different forms of data and documentation involved in an asylum case. As with other reclassification processes [following Bowker and Star 1999], credibility is an ambiguous prototypical concept for decision-makers to attempt certainty, especially important to consider in the design of data-driven technologies where stakeholders have differential power.