Efforts to augment accountability through the use of metrics, and especially randomised controlled trial or other statistical methods place an increased burden on small nongovernmental organisations (NGOs) doing global health. In this paper, we explore how one small NGO works to generate forms of accountability and evidence that may not conform to new metrics trends but nevertheless deserve attention and scrutiny for being effective, practical and reliable in the area of maternal and infant health. Through an analysis of one NGO and, in particular, its organisational and ethical principles for creating a network of safety for maternal and child health, we argue that alternative forms of (ac)counting like these might provide useful evidence of another kind of successful global health work.