Premise of the study: We developed a target enrichment panel for phylogenomic studies of Dioscorea, an economically important genus with incompletely resolved relationships. Methods: Our bait panel comprises 260 low-to-single-copy nuclear genes targeted to work in Dioscorea, and is assessed here using a preliminary taxon sampling that includes both distantly and closely related taxa, including several yam crops and potential crop wild relatives (CWRs). We applied coalescent-based and maximum-likelihood phylogenomic inference approaches to the pilot taxon set, incorporating new and published transcriptome data from additional species. Results: The custom panel retrieved ~94% of targets and >80% of full gene length from 88% and 68% of samples, respectively. It has minimal gene overlap with existing panels designed for angiosperm-wide studies, and generally recovers longer and more variable targets. Pilot phylogenomic analyses consistently resolve most deep and recent relationships with strong support across analyses, and point to revised relationships between the crop species D. alata and candidate CWRs. Discussion: Our customized panel reliably retrieves targeted loci from Dioscorea, is informative for resolving relationships in denser samplings, and suitable for refining our understanding of the independent origins of cultivated yam species; the panel likely has broader promise for phylogenomic studies across Dioscoreales.