<b><i>Background:</i></b> Atopic dermatitis (AD) is a chronic inflammatory skin disorder that is associated with higher rates of psychological disorders, but limited evidence supported the association with alexithymia, a psychoaffective dysfunction. <b><i>Objectives:</i></b> This study was aimed to investigate the occurrence of alexithymia in AD patients, compared to healthy subjects. <b><i>Methods:</i></b> This cross-sectional study assessed AD severity by the Eczema Area and Severity Index (EASI) score, sleeplessness and itch by a numeric rating scale (NRS), and alexithymia by the 20-item Toronto Alexithymia Scale (TAS-20) score. The association between disease characteristics and alexithymia was evaluated through several logistic regression models. <b><i>Results:</i></b> 202 AD patients and 240 healthy subjects were included in this study. The alexithymic personality trait (TAS-20 ≥51) was more frequently observed among AD patients compared to the control group (62.4% [126/202] vs. 29.2% [70/240], <i>p</i> &#x3c; 0.0001). In particular, alexithymia (TAS-20 score ≥61) was detected in a significantly higher number of AD patients than in the controls (27.7% [56/202] vs. 7.5% [18/240]; <i>p</i> &#x3c; 0.0001), whereas borderline alexithymia was detected in 34.6% (70/202) of AD patients compared to 21.7% of healthy controls. Alexithymia was more common among severe AD patients (43.6%) compared to mild AD patients (15.6%) and correlated with itch intensity and sleep disturbances. Among clinical variables, ordered logistic regression analyses revealed disease severity as predictor of alexithymia. Indeed, univariate analysis showed EASI score, sleep NRS, and itch NRS being significantly associated with alexithymia, while a multivariate model identified increased EASI score values as predicting factor. <b><i>Conclusion:</i></b>This study described alexithymia in AD patients correlating its occurrence with clinical AD severity markers (EASI score, itch, and sleeplessness) and identifying the increase in EASI score as predicting factor.