Generic Quality of Life (QoL) instruments are useful to compare different populations. However these instruments could be criticized for having some drawbacks, and disease-specific instruments could be preferred to measure those aspects of wellbeing influenced by a specific disease and to effectively measure health changes over the time. OBJECTIVES: We tested convergent validity and sensitivity to change over time of a generic and disease-specific pediatric questionnaires to evaluate wellbeing in children with Atopic Dermatitis (AD), a very frequent, chronic and disabling disease. METHODS: Data from the Costi-&-Outcomes-in-Dermatite-Atopica (CODA) naturalistic, prospective Cost-Of-Illness study, involving moderate and severe AD patients, were used. Patients aged 5-16 y.o. and/or their caregivers completed twice (at flare-up and after 2 months) the KINDL (children-reported version: KINDL-C, parent-reported version: KINDL-P; scores=0-100, higher score=higher QoL) and CDLQI (Children's-Dermatology-Life-Quality-Index, with scores=0-30, higher score=lower QoL). Patients' clinical status was evaluated with the SCORAD index (SCORing-Atopic-Dermatitis, possible score=0-100, higher score=higher severity). We tested convergent validity by investigating correlations between QoL instruments; sensitivity to change over time was tested with paired students' t tests, Standardized Response Mean (SRM), Effect Size (EF). RESULTS: Pediatric patients were 66, 43.9% male, median age=8.8 y.o., median SCORAD at enrolment=41.5 (3.0-85.0). CDLQI significantly correlated with KINDL-P (Spearman's r=-0.44 p=0.001) and KINDL-C (r=-0.36 p=0.008), KINDL-P sensitively correlated with KINDL-C (r=0.67 p<0.0001). At follow-up clinical severity significantly decreased (Student's paired t test, p<0.0001). Patients reported significant lower scores of CDLQI (Student's paired t test, p<0.05), while no statistical change was found with KINDL-P and KINDL-C. SRM and ES were moderate for CDLQI (SRM=0.44, ES=0.41) and low for KINDL-C (SRM=0.26, ES=0.26) and KINDL-P (SRM=0.12, ES=0.11). CONCLUSION: KINDL significantly correlated with CDLQI, anyway sensitivity to change results were moderate to low. Understanding these properties in QoL questionnaires is necessary to allow their appropriate use and interpretation of QoL data.