The translation of advances in cancer biology to drug discovery can be complicated by pharmacokinetic variation between individuals and within individuals, and this can result in unpredictable toxicity and variable antineoplastic effects. Previously unrecognized variables (such as genetic polymorphisms) are now known to have a significant impact on drug disposition. How can the pharmacokinetic variability of anticancer agents be reduced? This will require the understanding of correlations between pharmacokinetics and treatment outcomes, the identification of relevant patient parameters, mathematical modelling of individual and population pharmacokinetics, and the development of algorithms that will tailor doses to the individual patient.