Technical Abstract: Simulation models of management effects on weed demography and the experience of successful organic and low-external-input farmers point in the same direction: long-term weed management success depends on diversified strategies that attack multiple weed life stages. In an era of declining research budgets, modeling analyses can help target limited research funds and personnel at those empirical questions that are most critical. One area identified by both models and farmer groups as in need of greater research is weed seed bank ecology. Perturbation analysis of demographic models highlights seed bank persistence as the greatest driver of population growth rate in annual weed species, and farmer adages such as "One year's seeding, seven years weeding..." indicate the importance of reducing seed return. Yet few management tactics target weed seeds, and the factors controlling weed seedbank persistence remain largely unexplored. Ongoing work at the USDA-ARS Invasive Weed Management Unit in Urbana, IL, explores seed, soil and microbiological factors contributing to persistence of seeds in the soil seed bank. An ongoing dialog between empirical and theoretical approaches will aid further progress in this area.