The use of Genomic Selection has become popular due to its wide success in private and public breeding programs. The current dissertation aims to derive information that will be used to produce a single nucleotide polymorphisms (SNP) chip that can be used to successfully predict traits of economic importance for the North Dakota State University (NDSU) barley breeding program. Three training populations were tested to determine which should be used for predictive modeling. Multiple predictive models were employed to verify the most effective method for each individual trait. Through this research, four major findings were concluded: the successful identification of (i) the traits that are candidates for prediction, (ii) the most successful models for each trait, (iii) the minimum number of markers required to predict a trait, and (iv) the markers that should be included on NDSU barley breeding program’s new SNP chip for genomic selection.