ras and ras-like proteins are membrane-bound GTPases that function as key control switches in mitogenic signal transduction, cytoskeletal organization and membrane trafficking. Oncogenic mutants of ras genes are the transforming genes most frequently found in human cancers and ras has been one of the most intensively studied proteins of the past decade. It is now becoming clear that an expanding menagerie of other proteins interact with ras and influence its GTPase and guanine nucleotide exchange activities. For example, the gene for von Recklinghausen neurofibromatosis encodes a protein (neurofibromin) that is a GTPase-activating (GAP) protein for ras. For ras to function (or malfunction as the case may be), it must be located at its site of action on the inner surface of the plasma membrane. ras is targeted to and anchored to the membrane by virtue of an isoprenoid lipid group that is post-translationally attached to ras by a two-subunit enzyme called farnesyltransferase (FTase). Inhibitors of FTase are prime candidates for new drugs to antagonize ras action because they should even block consituitively-activated oncogenic variants of ras by preventing them from getting to the membrane. Through database searching and multiple sequence alignment, we have identified highly-conserved domains and sequence motifs in various families of proteins that interact with ras. For the rasGAP-related domain of neurofibromin in particular, these conserved regions have been used as targets for site- directed mutagenesis studies aimed at elucidating protein structure-function relationships and mechanisms of catalysis. For the FTase subunits in particular, conserved repetitive sequence motifs may account for protein- protein interactions and zinc-binding function. We also predicted that a new yeast gene, MAD2, involved in cell cycle control, is most probably a FTase alpha subunit. These predictions are currently undergoing experimental analysis.