Existing scholarship has documented the numerous methodological flaws as well as the social harms of hereditarianism. However, it has been less successful in explaining why such claims retain scientific credibility. I argue that this persistence is sustained by a structural process, which I call epistemic laundering: the acquisition and retention of scientific standing through circulation across settings governed by heterogeneous evaluative norms – including journals, conferences, and para-academic dissemination platforms – without a corresponding increase in the integrated warrant for treating the claims as reliable. I identify and analyse five laundering mechanisms: inferential-package fragmentation, category mobility, conceptual flipping, controversy laundering, and infrastructure bridging. I organise them into two conceptual clusters: evaluation-evasion mechanisms, which shape how assessment is distributed, deflected, or reallocated, and credibility-transfer mechanisms, which move and amplify credibility across institutional boundaries, converting fragmented local adequacy into the appearance of global scientific standing. I illustrate these mechanisms at two analytically connected scales drawn from intelligence research: the broader Human Diversity Foundation infrastructure, and the Pesta et al. (2020) meta-analysis in Intelligence. I conclude by examining editorial governance, the limits of refutation, and provenance disclosure as structural remedies.