The MMR – maternal mortality ratio – has risen from obscurity to become a major global health indicator, even appearing as an indicator of progress towards the global Sustainable Development Goals. This has happened despite intractable challenges relating to the measurement of maternal mortality. Even after three decades of measurement innovation, maternal mortality data are widely presumed to be of poor quality, or, as one leading measurement expert has put it, ‘guilty until proven innocent’. This paper explores how and why leading epidemiologists, demographers and statisticians have devoted the better part of the last three decades to producing ever more sophisticated and expensive surveys and mathematical models of globally comparable MMR estimates. The development of better metrics is publicly justified by the need to know which interventions save lives and at what cost. We show, however, that measurement experts’ work has also been driven by the need to secure political priority for safe motherhood and by donors’ need to justify and monitor the results of investment flows. We explore the many effects and consequences of this measurement work, including the eclipsing of attention to strengthening much-needed national health information systems. We analyse this measurement work in relation to broader political and economic changes affecting the global health field, not least the incursion of neoliberal, business-oriented donors such as the World Bank and the Bill and Melinda Gates Foundation whose institutional structures have introduced new forms of administrative oversight and accountability that depend on indicators.