BackgroundMolecular and genomic surveillance is becoming increasingly used to track malaria control and elimination efforts. Blood samples can be collected as whole blood and stored at − 20 °C until DNA extraction, or as dried blood spots (DBS), circumventing the need for a cold chain. Despite the wide use of either method, systematic comparisons of how the method of blood sample preservation affects the limit of detection (LOD) of molecular diagnosis and the proportion of DNA recovered for downstream applications are lacking.MethodsExtractions based on spin columns, magnetic beads, Tween-Chelex, and direct PCR without prior extraction were compared for whole blood and dried blood spots (DBS) using dilution series of Plasmodium falciparum culture samples. Extracted DNA was quantified by qPCR and droplet digital PCR (ddPCR).ResultsDNA recovery was 5- to 10-fold higher for whole blood compared to DBS, resulting in a 2- to 3-fold lower LOD for both extraction methods compared to DBS. For whole blood, a magnetic bead-based method resulted in a DNA recovery rate of 88–98% when extracting from whole blood compared to 17–33% for a spin-column based method. For extractions from DBS, the magnetic bead-based method resulted in 8–20% DNA recovery, while the spin-column based method resulted in only 2% DNA recovery. The Tween-Chelex method was superior to other methods with 15–21% DNA recovery, and even more sensitive than extractions from whole blood samples. The direct PCR method was found to have the lowest LOD overall for both, whole blood and DBS.ConclusionsPronounced differences in LOD and DNA yield need to be considered when comparing prevalence estimates based on molecular methods and when selecting sampling protocols for other molecular surveillance applications.