On-farm implementation of IWM includes the biological and economic rationalization of weed management decisions. This, in turn, requires the ability to predict how weed populations and pressures will change with farm management options such as tillage method. We have developed a set of algorithms that simulate seedling emergence of mixed populations of green and redroot pigweed on the basis of the ecophysiological responses of seed germination and shoot elongation to temperature and, for germination, to moisture. The algorithms were calibrated using field data from 1993 and evaluated with data from two locations in 1994 and 1995. Over both sites and years, for four tillage systems, cumulative emergence was predicted with an overall root mean square error of 1.7%. Overprediction in one year was attributed to moisture shortage. Errors were greater for moldboard plowed plots than for those no-tilled. This decreased ability to predict emergence with increased tillage (soil disturbance) suggests that the algorithms should be modified to account for increased heterogeneity in weed seed distribution and soil moisture within disturbed soil. The algorithms, which were explicitly designed for incorporation within a crop growth model (e.g., CROPSIM), could become a useful part of a decision support system to rationalize weed management. Importantly, they could help predict changes in weed populations as farm management is adjusted, thereby reducing economic and environmental risk in agroecosystems.