Technical Abstract: Cheatgrass (Bromus tectorum L.) has come to dominate millions of hectares of rangeland in the Intermountain western United States. Previous studies have hypothesized that one mechanism conferring a competitive advantage to this species is the ability to germinate rapidly at low temperatures in the fall, winter and spring and, therefore, initiate growth and establishment more rapidly than more desirable perennial bunchgrass species. In this experiment, we developed thermal-germination-response models for multiple seedlots of cheatgrass and five perennial grass species. We conducted sensitivity analysis on potential-cumulative-germination response to a 38-y simulation of field-variable conditions of seedbed temperature and moisture. Cheatgrass uniformly germinated at a higher rate, and required significantly less time to complete germination than any of the perennial species for all but the slowest seed subpopulations. A germination-rate-sum index was used to integrate relative response characteristics over time. This index showed that germination rate of the most rapidly-germinating subpopulations of cheatgrass were 2-5 times higher than for the other species tested. Model simulations of this type provide a more ecologically relevant basis for seedlot evaluation as they allow comparisons across a wide range of potential conditions that may be experienced in the field.