Objectives: Multisystem inflammatory syndrome in children (MIS-C) is a late manifestation of SARS-CoV-2 infection, which presents with symptoms ranging from milder mucocutaneous and gastrointestinal symptoms to severe cardiovascular and neurological manifestations. We studied the clinical, biochemical, hematological, and immunological characteristics of MIS-C patients to understand this disease entity and to identify markers of severe disease. Materials and Methods: Twenty-four MIS-C patients, four acute COVID-19 infections, and ten healthy controls (HC) from a tertiary care pediatric hospital in Mumbai were enrolled in the study. Clinical, biochemical, hematological, and immunological parameters comprising major lymphocyte, neutrophil, and monocyte subpopulations and key activator and inhibitory markers were studied. Statistical Analysis: All parameters were compared between the healthy, COVID and MIS-C groups at Day 0, 7 and 14 using non-parametric statistical tests. Machine learning tools were used for multivariate data analysis to identify the immunological parameters that could help predict severe disease. Results: NKp46pos NK cell (%), CD11 positive eosinophil (%), D-dimer, and Tim3pos Tc (%) were identified as the most important markers that could help predict severe disease, with NKp46pos NK cells as the top contributor. A disease severity metric utilizing these markers can be used to identify patients who are likely to have a severe course of disease. Conclusions: NK cells directly contribute to disease severity in MIS-C. As the JAK-STAT pathway is known to be important for NK cell development, maturation, and function, ruxolitinib, which is a JAK1/JAK2 inhibitor, might be beneficial in the management of this condition.