The current cure rate of 80% in childhood acute lymphoblastic leukaemia attests to the effectiveness of risk-directed therapy developed through well-designed clinical trials. In the past decade there have been remarkable advances in the definition of the molecular abnormalities involved in leukaemogenesis and drug resistance. These advances have led to the development of promising new therapeutic strategies, including agents targeted to the molecular lesions that cause leukaemia. The importance of host pharmacogenetics has also been recognised. Thus, genetic polymorphisms of certain enzymes have been linked with host susceptibility to the development of de novo leukaemia or therapy-related second cancers. Furthermore, recognition of inherited differences in the metabolism of antileukaemic agents has provided rational selection criteria for optimal drug dosages and scheduling. Treatment response assessed by measurements of submicroscopic leukaemia (minimal residual disease) has emerged as a powerful and independent prognostic indicator for gauging the intensity of therapy. Ultimately, treatment based on biological features of leukaemic cells, host genetics, and the amount of residual disease should improve cure rates further.