Wheat grain protein is an important quality determinant for hard spring wheat (HSW). Marker-assisted selection (MAS) may be a useful tool for plant breeders to identify high protein early generation material. The purpose of this study is to evaluate the effectiveness of using MAS in developing high grain protein content (GPC) spring wheat genotypes. Two populations were developed by crossing the cultivars ‘Ember' and ‘McVey' with ‘Glupro', a high GPC donor that contains a gene from a Triticum dicoccoides accession. Three methods were used to select 30 high GPC genotypes from each population cross; marker-assisted selection, near infra-red phenotypic selection, and a random control. Marker-assisted selection involved using the co-dominant microsatellite markers Xgwm193 and Xgwm508 on 250 F2 plants of each population. These markers flank the T. dicoccoides high GPC gene, and map over a 12.5 cM interval. Phenotypic selection was made from 250 F4 plants by selecting the highest 30 in GPC based on NIR readings. The random control involved selecting 30 genotypes from the F2 populations after which they were selfed to produce F4 lines. In 2004, the selected 30 genotypes from each population and method combination were grown in replicated field trials at three North Dakota locations. A contrast analysis between phenotypic NIR and MAS selection methods revealed that the mean GPC was greater among the 30 lines of both populations using phenotypic NIR selection at two locations and no significant difference was present at the third location. The random control had significantly lower mean GPC at all locations. In 2005 the experiment was repeated at three North Dakota locations. We will report on the effectiveness of each selection method based on genotype comparisons for GPC between methods for both years.