By incorporating demographic analyses of fitness components (e.g., survival and reproduction) within a reciprocal sowing design, we tested for 3 consecutive years whether local adaptation has occurred in the alien grass Bromus tectorum (cheatgrass) within 7 habitats along an environmental gradient from arid steppe to subalpine forest in the Intermontain Region of western North America. Patterns of emergence and survival were strongly influenced by the local environment. In terms of survival, expression of significant local adaptation in Tsuga heterophylla habitat varied among years. In contrast, relative differences in flowering time among seed sources were stable across sites and years. Populations from the arid steppe were the earliest to flower; flowering was latest in populations from the mesic Tsuga heterophylla habitat. In terms of net reproductive rate, evidence for local adaptation in B. tectorum was obtained in populations from habitats representing environmental extremes: an arid, saline site dominated by the shrub Sarcobatus vermiculatus and clearings within the cool, mesic Tsuga heterophylla forest habitat. Unlike the plants introduced from other sites, members of the resident population at the Sarcobatus site flowered and produced seeds before soil water became limiting. In contrast, net reproductive rates in other habitats were sometimes the lowest for populations in their home site. This lack of an advantage for local populations within more environmentally moderate sites suggests that limited dispersal may restrict the rate at which superior genotypes are introduced into a particular site.