In recent years, ‘vulnerability’ has been getting more traction in theoretical, professional and popular spaces as an alternative or complement to the concept of risk. As a group of science and technology studies scholars with different disciplinary orientations yet a shared concern with biomedicine, self and society, we investigate how vulnerability has become a salient and even dominant idiom for discussing disease and disease risk. We argue that this is at least partly due to an inherent indeterminacy in what ‘vulnerability’ means and does, both within and across different discourses. Through a review of feminist and disability theory, and a discussion of how vulnerability and disease both get recruited into a binary conceptualisation of normal versus abnormal, we argue that vulnerability’s indeterminacy is, in fact, its strength, and that it should be used differently than risk. Using COVID-19 management in the UK as an illustration of the current ambivalence and ambiguity in how vulnerability versus risk is applied, we suggest that instead of being codified or quantified, as it has started to be in some biomedical and public health applications, vulnerability and its remedies should be determined in conjunction with affected communities and in ways that are polyvalent, flexible and nuanced. The concept of vulnerability encapsulates an important precept: we must recognise inequality as undesirable while not attempting to ‘solve’ it in deterministic ways. Rather than becoming fixed into labels, unidirectional causalities or top-down universalising metrics, vulnerability could be used to insist on relational, context-specific understandings of disease and disease risk—in line with contemporary social justice movements that require non-hierarchical and non-universal approaches to problems and solutions.