. 2004. The influence of harvest management and fertilizer application on seasonal yield, crude protein concentration and N offtake of grasses in northeast Saskatchewan. Can. J. Plant Sci. 84: 205-212. The goal of pasture management is to match quantity and quality of herbage to the requirement of the grazing animal. The objective of this study was to examine the effects of harvest management and fertilizer application on annual and seasonal production, protein concentration and N-offtake by eight grass species adapted to the central Aspen Parkland. In this 3-yr splitsplit-plot experiment, grasses were either fertilized annually (50 and 13 kg ha -1 of N and P, respectively) or left unfertilized, and cut either twice annually (two-cut) in early July and late September, or three to four times annually (multi-cut) . Fertilizer increased yield of the two-cut treatment 15% (P < 0.05) more than the multi-cut treatment. On average, the multi-cut treatment had 32% higher protein concentration than the two-cut treatment throughout the season. Fertilizer significantly increased protein concentration only in the first harvest of the multi-cut clipped plots, probably because the application rate was low. Nitrogen offtake of the multi-cut treatment averaged 24% higher than the two-cut treatment, even though the latter yielded more. Meadow bromegrass showed the most uniform seasonal distribution of yield among the high-yielding grasses under all management systems while smooth bromegrass and crested wheatgrass were generally the poorest. Crested wheatgrass had significantly more first cut production than all other grasses. Meadow bromegrass herbage also had the lowest protein concentrations throughout the growing season among the high producing grasses. Since the protein concentrations were generally higher than required by grazing beef cattle, the meadow bromegrass herbage was the most suitable for this class of cattle. Green needlegrass also showed potential for use in pastures. . Les engrais augmentent le rendement du régime à double coupe de 15 % (P < 0,05) comparativement à celui du régime à coupe multiple. En moyenne, le régime à coupe multiple produit 32 % de protéines de plus que celui à coupe double durant la période végétative. Les amendements n'augmentent sensiblement la concentration de protéines qu'à la première récolte, sur les parcelles faisant l'objet d'une coupe multiple, sans doute à cause du faible taux d'application. Les parcelles à coupe multiple donnent en moyenne 24 % plus d'azote que celles qu'on récolte deux fois, malgré le rendement supérieur de ces dernières. Le brome des prés avait le rendement saisonnier le plus uniformément réparti des graminées à haut rendement, sans égard au système de gestion, alors que le brome inerme et l'agropyre à crête avaient généralement le rendement le plus mal réparti. L'agropyre à crête donne 205