Disability is sometimes theorised as existing between the world (including social norms and infrastructure) and the person (who gets labelled disabled in a ‘misfit’ between the world and them). Disability is often enacted through data systems and infrastructures and the history of disability studies reflects a fight against such systems. In this paper, we examine the fraught relationships between disability studies and data science — from institutions and historical marginalisation to current practices of policing and surveillance. We critique the resulting preeminence of qualitative methods in disability studies as one impediment to translating disability studies to data scientists and to effective policy-making. We then address hopeful movements to crip data studies, looking at work on AI and disability bias, crip technoscience, counterventional research, and cripped data.