Farnesyltransferase inhibitors (FTIs) are a novel class of cancer therapeutics whose development was based on the discovery that the function of oncogenic Ras depends upon its posttranslational farnesylation. Significantly, experiments in animal models have shown that FTIs have promise as nontoxic cancer therapeutics. However, cell biological studies have suggested that FTIs may act at a level beyond that of suppressing Ras function, so the exact mechanism of action has emerged as a question of major interest. Here, we review evidence that proteins other than Ras are important targets for inhibition, summarize findings suggesting a role for farnesylated Rho proteins prompted by studies on RhoB, and suggest a new model for how FTIs exert their biological effects. The 'FTI-Rho hypothesis' proposes that FTIs act in part by altering Rho-dependent cell adhesion signals which are linked to pathways controlling cell cycle and cell survival and which are subverted or defective in neoplastic cells. This model offers a novel framework for addressing the questions about FTI biology, including the basis for lack of toxicity to normal cells, cytotoxic versus cytostatic effects on tumor cells, and the persistence and drug resistance of malignant cells in FTI-treated animals.