The diagnosis and management of glucose-6-phosphate dehydrogenase (G6PD) deficiency is a crucial aspect in the current phases of malaria control and elimination, which will require the wider use of 8-aminoquinolines for both reducing Plasmodium falciparum transmission and achieving the radical cure of Plasmodium vivax. 8-aminoquinolines, such as primaquine, can induce severe haemolysis in G6PD-deficient individuals, potentially creating significant morbidity and undermining confidence in 8-aminoquinoline prescription. On the other hand, erring on the side of safety and excluding large numbers of people with unconfirmed G6PD deficiency from treatment with 8-aminoquinolines will diminish the impact of these drugs. Estimating the remaining G6PD enzyme activity is the most direct, accessible, and reliable assessment of the phenotype and remains the gold standard for the diagnosis of patients who could be harmed by the administration of primaquine. Genotyping seems an unambiguous technique, but its use is limited by cost and the large range of recognized G6PD genotypes. A number of enzyme activity assays diagnose G6PD deficiency, but they require a cold chain, specialized equipment, and laboratory skills. These assays are impractical for care delivery where most patients with malaria live. Improvements to the diagnosis of G6PD deficiency are required for the broader and safer use of 8-aminoquinolines to kill hypnozoites, while lower doses of primaquine may be safely used to kill gametocytes without testing. The discussions and conclusions of a workshop conducted in Incheon, Korea in May 2012 to review key knowledge gaps in G6PD deficiency are reported here.