Genomic selection (GS) is a recent development of marker-assisted selection. With GS the goal is to predict breeding values of selection candidates using a prediction model that is developed using a representative set of individuals that have both phenotypic and genome-wide marker data. Simulation and empirical studies have demonstrated that GS is one promising breeding tool that can help increase the rate of genetic gain for complex traits by enabling faster breeding cycles. Together with the Cornell University, we are investigating the efficiencies of various GS models for selecting for grain yield under irrigated system. We are also investigating the use of GS in biparental and multi parental populations for recurrent selection.