Brassica rapa is a model species for agronomic, ecological, evolutionary and translational studies. Here we describe high-density SNP discovery and genetic map construction for a Brassica rapa recombinant inbred line (RIL) population derived from field collected RNA-seq data. This high-density genotype data enables the detection and correction of putative genome mis-assemblies and accurate assignment of scaffold sequences to their likely genomic locations. These assembly improvements represent 7.1-8.0% of the annotated Brassica rapa genome. We demonstrate how using this new resource leads to a significant improvement for QTL analysis over the current low-density genetic map. Improvements are achieved by the increased mapping resolution and by having known genomic coordinates to anchor the markers for candidate gene discovery. These new molecular resources and improvements in the genome annotation will benefit the Brassicaceae genomics community and may help guide other communities in fine-tuning genome annotations.