There is compelling evidence that racial discrimination is a risk factor for illness and disease. But what are health scientists measuring—and what do they think they are measuring—when they include measures of racial discrimination in health research? We synthesize theoretical conceptualizations of racial discrimination in health research and critically assess whether and how these concepts correspond (or not) to widely used measures of racial discrimination. In doing so, we show that while researchers often use terms such as ‘self-reported discrimination', ‘perceptions of discrimination', and ‘exposure to discrimination' interchangeably, these concepts are indeed unique, with each holding a distinct epistemological position and theoretical and methodological capacity to uncover the impact of racial discrimination on health and health disparities. Importantly, we argue that commonly used measures of self-reported or perceived racial discrimination are just the ‘tip of the iceberg' in terms of revealing the ways in which discrimination shapes health inequities. Scientists and practitioners must be cognizant of and intentional in their measurement choices and language, as the framing of these processes will inform policy and intervention efforts aimed at eliminating discrimination.