The spatial distribution of the Pine Shoot Beetle (Tomicus piniperda L.) was studied in 1984 and 1985 in some stands of Scots pine (Pinus sylvestris L.) in the Orleans forest (France). The population density was estimated with the help of the number of shoots damaged by the insectsduring their summer maturation stage. These shoots were harvested and counted on a serie of transects. Assuming these results the beetles distribution in the reproduction and hivernation sites was approached. The estival distribution (fig. 1) appears relatively homogeneous on lar!c surfaces (0,6 to 1,0 shoot/m2), except some local concentrations, the spots (up to 20 shoots/m2 ). Two kinds of spots have been observed : one central spot, located in the center of forest compartment around a group of dying pines (fig. 2) ; two border spots, much more important, located at the border of a forest compartment, near to piles of cutting wood (fig. 6). Two artificial central spots have also been made by introducing infested logs in the center of a forest compartment. Around all these spots the density of the shoots decreases exponentially from the center to the periphery (fig. 5). The spatial distribution evolves during the year ; this is particularly clear around the spots (fig. 7). During the reproduction stage the insects are extremely aggregated and concentrated on the scarce places available for oviposition (weakened trees, logs...). During the maturation stage a slow dispersal of the insects is observed, which results from the successive attacks of shoots, the greatest dispersion state being reached at the hivernation times.In addition, from the number of harvested shoots and from experimental studies of the shoot consumption by T. piniperda, we have estimated and mapped the number of insects at the beginning of the maturation stage (fig. 8). The comparison of these results with the number of emergent callow beetles observed on the reproduction sites gives evidence of a systematic excess of emergent beetles into the spots and a systematic lack of emergent beetles out of these spots (tabl. 3). These results suggest that an important dispersal of the callow beetles takes place just after they get out of their reproduction sites. The dispersal flight would concern an average distance of about one kilometer. Consequently an exogenous population is observed out of the spots, slowly decreasing when the distance from the spots increases ; in this case, this population accounts for 40 or 80 p. 100 of the number of beetles maturing out of the spots. Then, the border spots have produced about two third of the maturing bark beetles located in 300 hectares. The two dispersal phenomena lead to the establishment of a minimum density of population all over the forested area before the spring swarming. In these conditions, the swarming beetles can easily find the available reproduction sites. Then, the large population changes are possibly an important element of the strategy of T. piniperda for exploiting temporary and unpredictable habitats which are the rule in endemic conditions.