Pharmacogenetics holds the promise of minimizing adverse neurodevelopmental outcomes of cancer patients by identifying patients at risk, enabling the individualization of treatment and the planning of close follow-up and early remediation. This review focuses first on methotrexate, a drug often implicated in neurotoxicity, especially when used in combination with brain irradiation. The second focus is on glucocorticoids that have been found to be linked to adverse developmental effects in relation with the psychosocial environment. For both examples, we review how polymorphisms of genes encoding enzymes involved in specific mechanisms of action could moderate adverse neurodevelopmental consequences, eventually through common final pathways such as oxidative stress. We discuss a multiple hit model and possible strategies required to rise to the challenge of this integrative research.