Kelp aquaculture has rapidly expanded over the past decade, driven by its sustainability, restorative potential, and nutritional value. The United States native Saccharina latissima (sugar kelp) and Macrocystis pyrifera (giant kelp) are promising species for cultivation efforts to support growing food demands amid environmental pressures on conventional food sources. The integration of genetic tools into breeding programs has been demonstrated to accelerate crop improvement and allow for more precise trait selection, and the recent release of an annotated genome assembly for North American S. latissima provides a valuable resource for genomics-assisted breeding in kelp. A critical goal for the future of kelp cultivation is the development of technologies to control kelp reproduction to address environmental concerns of gene flow. Utilizing the genomes of these kelps, we analyzed whole-genome sequencing (WGS) data from 278 S. latissima and 512 M. pyrifera haploid gametophytes to predict potentially non-reproductive crosses from genetic variants. We pre-screened candidate loci linked to reproduction for variants with a predicted high degree of impact on protein products (e.g., nonsense and missense mutations). We predicted putative non-reproductive test crosses from the set of genotyped gametophytes. For S. latissima, several crosses showed predicted infertility across multiple seasons. However, predictions for M. pyrifera are still undergoing experimental validation. Our findings demonstrate the utility of genomic resources in kelp trait prediction, paving the way for more economical and targeted breeding initiatives in marine aquaculture.