In October, 2021, WHO recommended that the RTS,S malaria vaccine, with its strong safety profile and high impact, be provided to children from age 5 months in regions with moderate to high Plasmodium falciparum malaria transmission. The evidence base included phase 3 trials in seven African countries and an ongoing malaria vaccine implementation programme (MVIP) in three African countries. We highlight problems with the MVIP mortality data, including potential confounding, inappropriate use of severe malaria as a surrogate marker, a statistically non-significant effect, and assessment after 2 years instead of the stipulated 4 years, which could have inflated the benefits and deflated the risks associated with the vaccine. We conclude that the claimed impact of the MVIP on mortality is not based on enough scientific evidence and that the MVIP findings do not rule out the possibility of increased mortality among vaccinated girls compared with vaccinated boys, as observed in the phase 3 studies. The MVIP should adhere fully to the planned analyses and the data should be made available for independent assessment. Roll-out of the vaccine elsewhere should include rigorous evaluation, especially of its safety.