Population genomics is a useful tool in the integrated pest management toolbox for elucidating population dynamics, demography, and histories of invasion. However, next-generation sequencing approaches can be hampered by low DNA input from small organisms, such as insect pests. Here, we use a restriction-site associated DNA sequencing approach combined with whole-genome amplification to assess genomic population structure of a newly described pest of canola, the diminutive canola flower midge, Contarinia brassicola. We find that whole-genome amplification prior to library preparation caused a reduction in the overall number of loci sequenced and an increase in overall sequencing depth but had no discernable impact on genotyping consistency for population genetic analysis. Clustering analyses recovered little geographic structure across the main canola production region, but differentiated several geographically disparate populations at edges of the agricultural zone. Given a lack of alternative hypotheses for this pattern, we suggest these data support alternative hosts for this species and thus our canola-centric view of this midge as a pest has limited our understanding of its biology. These results speak to the need for increased surveying effort across multiple habitats and other potential hosts within Brassicaceae, to elucidate both our ecological and evolutionary knowledge of this species as well as potential management implications.