Most anticancer drugs are characterised by a narrow therapeutic window; hence, a small change in dose can lead to poor antitumour effects or an unacceptable degree of toxicity. The rationale for using body surface area (BSA) to dose antineoplastic agents is to normalise the effects of drugs, and accordingly, it has been routinely employed as the only independent variable. In the last 10 years, however, several studies have shown a poor relationship of BSA for predicting drug exposure, and an irrelevant correlation between this variable and pharmacokinetic (PK) parameters. In this paper, the results of this relationship for various commonly used antineoplastic agents are reviewed, and the influence of BSA to decrease the total variability in clearance among patients is underlined. As reported, BSA failed to individualise the effects of the majority of the agents explored. The criteria that can predict a clinically meaningful relationship between BSA and drug clearance are discussed, and some alternative strategies to dose agents when BSA has proven to be useless are proposed.