Atopic dermatitis (AD) is one of the most prevalent chronic inflammatory dermatological disorders in childhood. Assessment of AD severity is the initial step in designing the proper therapeutic plan. Moreover, it is imperative for evaluation of disease improvement during and following therapy. This study was designed to assess the prognostic role of miRNA-155 (miR-155) in the prediction of AD severity as the primary outcome. While the secondary outcome was to correlate the serum miR-155 expression levels with the scoring atopic dermatitis (SCORAD) severity index. This case-control study included 24 children with AD and 24 apparently healthy children as a control group. AD children were stratified according to the SCORAD severity index. Approximately 58% of children had mild AD, 25% moderate AD, and about 17% severe AD. Children with AD had statistically significantly higher miR-155 expression levels in comparison to the control children, (p< 0.001). Children with severe AD had statistically significantly higher miR-155 expression levels compared to mild AD children (p=0.001). The receiver operating characteristic curve analysis for miR-155 demonstrated that miR-155 can differentiate between children with mild AD and those with moderate-to-severe AD, with an area under the curve of 0.879, and an excellent discrimination power. A statistically strong significant positive correlation existed between miR-155 levels and SCORAD severity index (rs= 0.666, p<0.001). In conclusion, MiR-155 could be considered as a non-invasive biomarker of AD severity in children. It is a promising prognostic tool in the prediction of AD severity.