Evans syndrome (ES) is a rare severe autoimmune disorder characterized by the combination of autoimmune hemolytic anemia and immune thrombocytopenia. In most cases, the underlying cause is unknown. We sought to identify genetic defects in pediatric ES (pES), based on a hypothesis of strong genetic determinism. In a national, prospective cohort of 203 patients with early-onset ES (median [range] age at last follow-up: 16.3 years ([1.2-41.0 years]) initiated in 2004, 80 nonselected consecutive individuals underwent genetic testing. The clinical data were analyzed as a function of the genetic findings. Fifty-two patients (65%) received a genetic diagnosis (the M+ group): 49 carried germline mutations and 3 carried somatic variants. Thirty-two (40%) had pathogenic mutations in 1 of 9 genes known to be involved in primary immunodeficiencies (TNFRSF6, CTLA4, STAT3, PIK3CD, CBL, ADAR1, LRBA, RAG1, and KRAS), whereas 20 patients (25%) carried probable pathogenic variants in 16 genes that had not previously been reported in the context of autoimmune disease. Lastly, no genetic abnormalities were found in the remaining 28 patients (35%, the M− group). The M+ group displayed more severe disease than the M− group, with a greater frequency of additional immunopathologic manifestations and a greater median number of lines of treatment. Six patients (all from the M+ group) died during the study. In conclusion, pES was potentially genetically determined in at least 65% of cases. Systematic, wide-ranging genetic screening should be offered in pES; the genetic findings have prognostic significance and may guide the choice of a targeted treatment.