AbstractObjectiveAnxiety symptoms are one of the most frequent manifestations in people attending primary care, although how the symptoms are associated is unclear. This study aimed to establish the symptom structure of the Generalized Anxiety Disorder scale (GAD‐7) using a novel network approach in combination with traditional analytical tools.MethodsA sample of 1704 primary care patients with emotional disorders (i.e., anxiety, depression, and/or somatization) completed the GAD‐7 to report their anxiety symptoms. We examined the GAD‐7 structure using exploratory graph analysis (EGA) compared to exploratory factor analysis (EFA) and confirmatory factor analysis.ResultsThe EFA results showed a one‐factor solution, but EGA revealed a two‐factor solution (cognitive–emotional and somatic). “Worrying too much” and “difficulty relaxing” were the most relevant symptoms.ConclusionsThe results support the possible distinction between the somatic and cognitive–emotional components of the GAD‐7, thus permitting more specific screening in primary care settings.