Perennial weeds have become an increasing problem with the reduction of herbicide use and tillage. Because of their capacity to regenerate from vegetative organs, such as roots and rhizomes, they must be controlled with different management techniques than annuals. To date, few studies exist on this topic and farmers often lack solutions to control perennials in their cropping systems. Management strategies must focus on the depletion of the regenerative organ reserves to hamper the production of new shoots. However, there is a need for a better understanding of below-ground organs physiology and the determining factors of their regrowth capacities. Our aim was to assess the number of shoots and their growth speed from one below-ground fragment according to 1) its weight, which is related to the amount of available reserves, and 2) the number of buds initially present (visible) on it. A pot experiment was carried out in a greenhouse: fragments of rhizomes of Elymus repens with different weights and different visible bud numbers were buried into 6-L pots. The number and length of the produced shoots were measured dynamically during the experiment. The effects of fragment weight and bud number on (1) the probability that a fragment produce at least one shoot and (2) shoot growth speed. The main results indicate a strong positive impact of the fragment weight, meaning that the more the reserves, the faster shoots grow, and a negative impact of the density of shoots on a same fragment, which implies a competition effet between the growing shoots. More perennial species are to be studied with the same intentions. These results will be used to model regrowth of perennials from below-ground fragments after tillage in a cropping system simulation model (FLORSYS). This model simulates daily weed dynamics and crop production over the years from cropping system and pedoclimate. The ultimate goal will be to identify agroecological weed management strategies and to help farmers to redesign their cropping systems.