BACKGROUND: Atopic dermatitis (AD) is a common inflammatory disease with heterogeneous clinical features. Certain meaningful phenotypes and clinical features may help better classify AD patients for personalized medicine. To our knowledge, no ideal predictors have been found so far. We aim to investigate clinical predictors for the 4-week efficacy of dupilumab treatment in AD patients in the real world. METHODS: Two hundred and thirty-three AD patients treated with dupilumab were enrolled between June 2020 and December 2023. Patients' information, characteristics, and medical history were collected. They were evaluated at the baseline and 4 weeks after dupilumab treatment and divided into groups according to the Investigator's Global Assessment (IGA) and peak pruritus numerical rating scale (PP-NRS) score. Statistical analyses were used to evaluate potential predictors. RESULTS: Age increase, late-onset, erythroderma type, and elevation of total serum IgE level were risk factors for poor response after 4-week treatment. Female, atopic personal or family history, and allergic rhinitis were factors for better response. Multivariate logistic regression analysis showed extrinsic AD had a poor reaction than intrinsic AD (OR=4.792,95% CI 1.460-15.732, CONCLUSION: Extrinsic AD and chronic lichenoid eczema are risk factors for poor response to 4-week dupilumab treatment, which suggests that AD patients with extrinsic and chronic lichenoid eczema may need more patience for long-term treatment or make other options.