Abstract : Obesity currently represents a major public health issue. Despite the fact that physical activity (PA) is recognized to play a key role in weight control, numerous studies tend to show that obese populations experience difficulties in the adoption of an active lifestyle. This doctoral dissertation deals with the psychosocial determinants of PA among obese populations. A meta-analysis was first conducted so as to determine the impact of interventions promoting PA among obese populations. Inclusion of 49 studies published to date enabled to conclude to the existence of a significant, but moderated (d=0.44), impact of interventions on participants' PA. Despite the fact that few variables (e.g., duration, type of PA indicator, format of sessions) modulated impact of interventions, few results appeared concerning the psychosocial processes implicated in those effects. Based on self-determination theory (SDT, Deci & Ryan, 2002), a program of research comprising 3 complementary studies was then conducted in order to (1) enhance the understanding of the motivational mechanisms associated with a PA adoption among obese adolescents, and (2) evaluate the impact of a motivational intervention aiming at helping this population to become more physically active. Using a cross-sectional methodology, study 1 aimed at indentifying the motivational patterns specifically associated with different type of PA (i.e., leisure time PA, sport, daily PA). The results suggest that leisure time PA is associated with high levels of identified and integrated regulations, as well as low levels of introjection. Sport participation was associated with high levels of introjection and low levels of external regulation. Study 2 explored with a qualitative methodology the motivational factors associated with the success or failure of a weight loss program for adolescents. The results indicate that participants who succeeded in losing weight expressed more support from their family and more autonomous reasons to maintain health behaviors. Lastly, study 3 consisted in a randomized controlled trial of 6 months aimed at determining to what extent the addition 6 motivational interviewing sessions (MI, Miller & Rollnick, 2002) would impact the efficacy of a cognitive-behavioral weight loss program. Compared to participants who only benefited from the weight loss program, multilevel analyses reveal that adolescents who received both MI and weight loss program reported (1) an higher increase of PA (self-reported and objective), (2) an higher increase of integrated and identified regulations, (3) an higher autonomy support from medical staff and (4) a more important decrease of amotivation. Taken globally, those results both suggest interest of MI as a pertinent counseling method to promote PA among obese adolescent population as well as interest of SDT as an explanatory framework of motivational processes associated with PA adoption among obese adolescents.