Résumé : Attention-deficit/hyperactivity disorder (ADHD) is characterized by executive dysfunction and difficulties in processing emotional facial expressions, yet the large-scale neural dynamics underlying these impairments remain insufficiently understood. This study applied network-based EEG source analysis to examine oscillatory cortical activity during cognitive and emotional Go/NoGo tasks in individuals with ADHD. EEG data from 272 participants (ADHD n equals 102, controls n equals 170, age range 6 to 60 years) were analyzed using exact low-resolution brain electromagnetic tomography combined with functional independent component analysis, yielding ten frequency-resolved cortical networks. Mixed-effects ANCOVAs were conducted on independent component loadings with Group, Task, and Condition as factors and age and sex as covariates. ADHD participants showed statistically significant but small increases in activation across several networks, including a gamma-dominant inferior temporal component showing a Group effect and a Group by Condition interaction with stronger NoGo-related activation in ADHD. Two additional components showed similar but weaker NoGo-selective patterns. A main effect of Task emerged only for one temporal delta component, with higher activation during the VCPT than the ECPT. No Group by Task interactions were observed. Behavioral results replicated the established ADHD performance profile, with slower responses, greater variability, and higher error rates, particularly during the emotional ECPT. Overall, the findings reveal subtle alterations in oscillatory brain networks during inhibitory processing in ADHD, with modest effect sizes embedded within substantial within-group variability. These results support a dimensional view of ADHD neurobiology and highlight the limited discriminative power of network-level EEG markers.