Patients vary widely in their responses to drug therapy, often manifested as adverse drug reactions, drug resistance or drug-associated toxicity. Functional polymorphisms in genes encoding enzymes, which are involved in drug action, may underlie these differences thus offering a powerful tool in predicting a treatment outcome. Several studies have addressed the potential for tailoring individual acute lymphoblastic leukemia (ALL) therapy based on patients genetics. Several candidate genes have been shown to have a predictive role, among which the best examples are the well-characterized polymorphisms in the thiopurine methyltransferase (TPMT) gene. The impact of TPMT genotypes on 6-MP tolerance, affecting the duration of treatment and the appearance of severe hematotoxicity or secondary malignancies has been well documented. Recent studies suggest that polymorphisms in enzymes of the folate cycle may modify the therapeutic effectiveness and toxicity of antifolates. Polymorphism in the enhancer element located in 5-UTR of thymidylate synthase gene influenced the outcome of ALL, whereas variants of methylene tetrahydrofolate reductase gene correlated with methotrexate sensitivity. Efforts have been also made to gain a closer insight into the role of polymorphisms in genes that might affect both disease susceptibility and drug metabolism, and some of them (e.g. glutathione S-transferase or quinone oxidoreductase) seem to affect the risk of recurrent disease in children with ALL. Extending the pharmacogenetics concept to other candidate genes / enzymes and to other drugs might, through comprehensive genetic evaluation, help clinical management of ALL patients.