To predict weed emergence and help farmers make weed management decisions, we constructed a mathematical model of seed germination for green and redroot pigweed based on temperature and water potential (moisture) and expressing cumulative germination in terms of thermal time (degree days). Empirical observations indicated green pigweed germinated at a lower base temperature than redroot pigweed but the germination rate of redroot pigweed is much faster as mean temperature increases. Moisture limitation delayed seed germination until 23.8 C (green pigweed) or 27.9 (redroot pigweed); thereafter, germination was independent of water potential as mean temperatures approached germination optima. Our germination model, based on a cumulative normal distribution function, accounted for 80 to 95% of the variation in seed germination and accurately predicted that redroot pigweed would have a faster germination rate than green pigweed. However, the model predicted that redroot pigweed would germinate before green pigweed (in thermal time) and was generally less accurate during the early period of seed germination. The model also predicted that moisture limitation would increase, rather than delay, seed germination. These errors were related to the mathematical function chosen and analyses used, but an explicit interaction term for water potential and temperature is also needed to produce an accurate model. We also tested the effect of mean temperature on shoot elongation (emergence) and described the relationship by a linear model. Base temperatures for shoot elongation were higher than for seed germination. Shoot elongation began at 15.6 and 14.4 C for green and redroot pigweed, respectively; they increased linearly with temperature until the optimum of 27.9 C was reached. Elongation was dependent on completion of the rate-limiting step of radicle emergence and was sensitive to temperature but not moisture; hence, elongation was sensitive to a much smaller temperature range. Beyond mathematical changes, we are testing our model in the field and need to link it to ecophysiological, genetic, and spatially explicit population processes for it to be useful in decision support for weed management.