Bramble (Rubus fruticosus L.) with many subspecies and very broad ecological amplitude is a common weed in temperate forest understorey and also in other world ecosystems where it is recorded as an invasive plant. It has been described to efficiently compete for water and nutrients and consequently to be able to compromise tree regeneration. However little is known on its development relative to light and whether this species could be controlled by tree cover to limit its influence on light attenuation in the understorey. We described bramble canopy architecture, cover relative to light and light interception in 60 bramble bushes from 18 sites in France and England with contrasted soils and climates. Canopy architecture was assessed by direct measurements of leaf area, cane length, basal diameter and 3D-digitizing. Its cover was quantified by vertically projected points using rods. Light beneath the tree canopy was measured using PAR sensors (Solem), whereas light interception by bramble bushes was measured using a PAR-ceptometer. Total bush leaf area, leaf number and mean leaf area were not or slightly affected by the site and soil conditions. At a cane scale, cane leaf area and leaf number were predicted with a good accuracy by the cane length; at a leaf scale, leaf area and leaflet areas were deduced from leaf length. The site had a significant influence on the slope of the relations. Leaf inclination distribution was planophile. Following the Beer-Lambert law, a good relation was established between light interception and LAI of the different bramble bushes, with no significant site influence, meaning that a single law can be used to predict light interception by bramble in different conditions. The LAI can be deduced either by the allometric relationships, either by visual estimation of canopy cover after adjustment with the rod technique. Bramble LAI and cover increased logarithmically with light availability to reach 30% cover at only 5 to 7 % light. Consequently bramble is able to tolerate deep shade, which is not in accordance with its reputed moderate to high light requirement reported in the literature. It seems thus difficult to control this species by tree canopy manipulations.