Abstract Much of our food and services rely on the work of migrants, which is often carried out under exploitative terms. Exploitative work results in work-related injuries and illnesses and mental ill-health. Yet, the global health community has been negligent in detecting migrant workers in the general population, and in measuring the effects of their work terms on their health. For instance, the International Labour Organization’s global labour force and forced labour surveys capture few health data. Conversely, health surveys, such as the Demographics and Health Survey (DHS), include few questions on exploitative and hazardous work conditions or aspects of migration status. Studies are even more remiss when it comes to developing evidence for work-related health interventions. This presentation will draw on two recent mixed-methods studies - on cattle ranching in the Amazon region of Brazil (Pará State) (n = 1,241), and on child domestic work in Liberia (n = 1,088) and Nigeria (n = 1,088). Findings suggest some necessary improvements to the measurement of migrant labour exploitation and its health effects, and approaches to gathering intervention-focused evidence. Each study highlights the value of re-thinking data collection beyond disciplinary silos and narrow dichotomous measures (e.g., forced/unforced labour), towards generating a more specific understanding of the different dimensions of labour exploitation (e.g., pay conditions, hours, occupational health and safety training and PPE), various types of associated harm, and how they compare across sectors and contexts. Further, each study demonstrates the benefits of intervention-focused research methods that are designed to feed directly into participatory intervention development processes. On this basis, we will suggest shifts in measurement and research designs to foster operationalizable evidence on migrant workers’ working conditions, health, and worker-informed interventions.