AbstractProducers on the Canadian Prairies grow barley (Hordeum vulgare L.) as a forage crop for ruminants, especially cattle. The purpose of our study was to determine methods to improve our selection of barley lines with superior forage value for ruminants. Data from the Western Cooperative Forage Barley Registration Test were used with permission of the Prairie Recommending Committee for Oat and Barley (PRCOB) and the breeders. Data were analyzed from 20 environments collected from 3 yr (2012–2014) and seven sites. We measured forage accumulation (FA) at the soft dough stage, five nutritive value traits, and three derived cow–calf variables. Although the variation in FA and forage nutritive value traits attributed to environment (E) was large, significant genotypic (G) effects were found for all traits measured. Whereas G× E effects were also significant, the mean square effect for G was up to 40 times higher than the mean squares for G×E. The proportion of variance attributed to G was greater for nutritive value traits than for FA. Significance of G means there was variability in the germplasm tested, and improvement through selection should be possible. The two parametric methods of stability analyses (Francis–Kannenberg method and Eberhart–Russell method) often identified the same lines as superior. Lines that were superior for FA and carrying capacity (cow on swath grazing, CC) were often different from those identified as superior for nutritive value, empty body weight gain (cow, EBWG) and average daily gain (backgrounding calf, ADG). A nonparametric method used to determine dynamic stability (Nassar–Hühn method) identified different lines as superior from the parametric methods. Breeders can use a combination of these selection criteria in breeding forage barley varieties with enhanced forage value for ruminants.