Juvenile myelomonocytic leukemia (JMML) is a rare clonal myelodysplastic/myeloproliferative disorder that affects young children. It is characterized by hypersensitivity of JMML cells to granulocyte-macrophage colony-stimulating factor (GM-CSF) in vitro. The pathogenesis of JMML seems to arise from constitutional activation of the GM-CSF/RAS (a GTPase) signaling pathway, a result of mutations in RAS, NF1, PTPN11, and CBL that interfere with downstream components of the pathway. Most patients with JMML usually experience an aggressive clinical course, and hematopoietic stem cell transplantation (HSCT) is currently the only curative treatment, although the high rates of relapses and graft failures are of great concern. In contrast, a certain proportion of patients experience a stable clinical course for a considerable period of time, and sometimes the disease even spontaneously resolves without any treatment. Recent studies have provided us with increased knowledge of genotype-phenotype correlations in JMML, and suggested that differences in clinical courses may reflect genetic status. Thus, genotype-based management is of current international interest, especially for JMML with RAS mutations. Cumulative evidence suggests that RAS mutations can be related to favorable clinical outcomes, and HSCT may not have to be a mandatory therapeutic option for a portion of patients with this mutation, although a consensus regarding genotype-based management has not yet been achieved. Further efforts toward identifying which patients who will do well without HSCT are required.