Pre-harvest aflatoxin contamination (PAC) occurs when the peanut is exposed to severe drought and heat stress prior to harvest. Development of genotypes that are either resistant or limiting to Aspergillus infection and/or that can curtail aflatoxin production is a key objective in peanut breeding. Using SNP markers and phenotype data on aflatoxin resistance collected in previous years, we utilized genomic selection (GS) to study PAC resistance in peanut. The technique was validated in the Florida-07 x GP-NC WS 16 recombinant inbred population by assessing high heritability traits. GS was then deployed to study PAC resistance in the Tifrunner x C76-16 and Florida-07 x C76-16 populations. The resultant models yielded prediction accuracy values of 0.24 and 0.23 which while low, were comparable to the heritability values of 0.31 and 0.1 for each population, respectively. Using genomic estimated breeding values (GEBVs), entries were selected for two drought shelter studies. Manual scoring and multispectral imaging were used to acquire end season drought stress data. Low correlation values (-0.15 for shelter A and 0.32 for shelter B) were observed between the GEBVs and actual end season aflatoxin content. Correlations between visual drought ratings and PAC were also low (0.23 and 0.16 for shelter A and B respectively). While strong inverse correlations were observed between GNDVI and visual drought ratings (-0.74 and -0.73 for shelters A and B respectively), GNDVI did not confer a clear advantage for selection of PAC resistance. Ideally, the use of genome-spanning markers in GS may enable selection for a difficult to phenotype trait like PAC resistance. Although this study did not show clear advantages of the method over conventional selection, it is an important step for implementation of GS and the use of GEBVs for trait selection in peanut.