Hypusine formation on the eukaryotic initiation factor 5A (eIF-5A) precursor represents a unique posttranslational modification that is ubiquitously present in eukaryotic cells and archaebacteria. Specific inhibition of deoxyhypusine synthase leads to growth arrest and cell death. The precise cellular function of eIF-5A and the physiological significance of hypusine modification are not clear. Although the methionyl-puromycin synthesis has been suggested to be the functional assay for eIF-5A activity in vitro, the role of eIF-5A in protein synthesis has not been established. Recent studies have suggested that eIF-5A may be the cellular target of the human immunodeficiency virus type 1 Rev and human T cell leukemia virus type 1 Rex proteins. Motif analysis suggested that eIF-5A resembles a bimodular RNA-binding protein in that it contains a stretch of basic amino acids clustered at the N-terminal region and a leucine-rich stretch at the C-terminal region. Using Rev target RNA, RRE, as a model, we tested the hypothesis that eIF-5A may be an RNA-binding protein. We found that both deoxyhypusine and hypusine-containing eIF-5A can bind to the 252-nt RRE RNA, as determined by a gel mobility shift assay. In contrast, the unmodified eIF-5A precursor cannot. Deoxyhypusine-containing eIF-5A, but not its precursor, could also cause supershift of the Rev stem-loop IIB RRE complex. Preliminary studies also indicated that eIF-5A can bind to RNA such as U6 snRNA and that deoxyhypusine modification appears to be required for the binding. The ability of eIF-5A to directly interact with RNA suggests that deoxyhypusine formation of eIF-5A may be related to its role in RNA processing and protein synthesis. Our study also suggests the possibility of using a gel mobility shift assay for eIF-5A-RNA binding as a functional assay for deoxyhypusine and hypusine formation.