Population screening follows the logic of secondary prevention: a population is screened to detect disease early and to initiate treatment before symptoms emerge. However, not all population screening is justifiable under all circumstances. In this article, we unpack Wilson and Jungner’s requirement that knowledge about the natural history of a disease must be ‘adequate’ for screening to proceed. We argue that any prior understanding of disease is inevitably found to be insufficient once population screening is instituted. Drawing upon ethnographic observations of clinical consultations and staff meetings conducted in a California regional clinical centre for metabolic-genetic disorders, we introduce the notion of bridging work to draw attention to the collective activities of the genetics team to revise the ontological nature of conditions unsettled by population-based newborn screening. Bridging work refers to the many activities required to reconcile the promise of technologies with the realities of their implementation. We illustrate how clinicians bridge the gap between what was known about a disease prior to screening and anomalous screening results, leading to an ontological transformation of disease categories.