ABSTRACT During the recovery period, athletes present high sedentary behaviour (SB). In non-athletes, there is a direct relationship between SB and obesity. However, little is known about this relationship in athletes. We hypothesized that different types of SB entail different associations with body composition outcomes. We examined the associations between different types of SB and body composition outcomes in 135 athletes (70 males) aged 21.3 ± 3.9 years old. Dual-energy X-ray absorptiometry was performed to assess fat mass (FM), fat-free mass (FFM) and trunk fat mass (TFM). A validated SB questionnaire (PACE) was used to estimate total SB and specific sedentary pursuits on an average day. Multiple linear regression analyses were performed, adjusting for age, sex, weekly training time, years of sport practice, and sport type. A positive association was found for total SB and total screen time with %TFM (β = 0.220, p = 0.038 and β = 0.319, p = 0.040, respectively), while an inverse association was found for %FFM (β = −0.214, p = 0.042 and β = −0.310, p = 0.026). A higher total screen time was related with a higher %FM (β = 0.283, p = 0.035). With a much stronger effect size, cell phone screen time was positively related with %FM (β = 1.447, p = 0.001). There is still debate whether high levels of physical activity can fully counteract the harms of SB. These findings suggest that SB, particularly cell phone screen time, can compromise athletes’ body composition, regardless of weekly training time. Sport federations and coaches may improve athletes’ body composition by targeting specific sedentary pursuits, i.e. total screen time and cell phone screen time, during athletes’ recovery time. Highlights Even in athletes, SB can compromise total and regional body composition regardless of high training time. Screen time, specifically cell phone screen time seems to be the sedentary pursuit to target in athletes. Sport federations and coaches must control SB during athletes' recovery time.