Abstract Objectives To stratify adult-onset Still’s disease (AOSD) patients in distinct clinical subsets to be differently managed, by using a multi-dimensional characterization. Methods AOSD patients were evaluated by using a hierarchical unsupervised cluster analysis comprising age, laboratory markers systemic score and outcomes. The squared Euclidean distances between each pair of patients were calculated and put into a distance matrix, which served as the input clustering algorithm. Derived clusters were descriptively analysed for any possible difference. Results Four AOSD patients clusters were identified. Disease onset in cluster 1 was characterized by fever (100%), skin rash (92%) and arthritis (83%), with the highest ferritin levels [mean (S.D.) 14 724 (6837) ng/ml]. In cluster 2, the onset was characterized by fever (100%), arthritis (100%) and liver involvement (90%), together with the highest CRP levels [288.10 (46.01) mg/l]. The patients in cluster 3 presented with fever (100%), myalgia (96%) and sore throat (92%). The highest systemic score values [8.88 (1.70)] and the highest mortality rate (54.2%) defined cluster 3. Fever (100%) and arthritis (90%) were the symptoms at the onset in cluster 4, which was characterized by the lowest ferritin and CRP levels [1457 (1298) ng/ml and 54.98 (48.67) mg/l, respectively]. Conclusion Four distinct phenotypic subgroups in AOSD could be suggested, possibly associated with different genetic background and pathogenic mechanisms. Our results could provide the basis for a precision medicine approach in AOSD in an attempt to find a clinical and laboratory multidimensional stratification and characterization, which would drive a tailored therapeutic approach in these patients.