Fungal diseases in cotton, such as various types of root rot and wilt symptoms, have been known for over 100 years, causing destruction of a large percentage of the cotton crop and large economic loss throughout the cottonbelt. It is now well known that almost all plants naturally produce their own defensive proteins, called pathogenesis-related (PR) proteins, directed against pathogens such as fungi and bacteria. PR proteins called osmotins are made in response to fungal pathogen stress or osmotic stress (water deprivation or salt exposure). Osmotins are thought to actually form pores in fungal membranes, leading to osmotic rupture and destruction of the fungal cells. Tobacco and potato osmotin genes have been shown to have biological control regions for overproduction of osmotin when the plants are exposed to fungal pathogens or drought. Since osmotins have antifungal activity against most fungal species, it should be possible to use the osmotin gene as a novel defense gene effective against numerous pathogenic fungi. A cotton genomic library harbored in lambda phage was screened with a tobacco osmotin hybridization probe to isolate prospective osmotin genes, in order to study their regulation of gene expression. A cotton genomic clone containing a 16.0-kb DNA segment was found to encompass two osmotin genes. One gene is full-length with an open reading frame of 729 basepairs without any introns, and encodes a presumptive osmotin-like preprotein of 242 amino acids. The other gene is almost full-length, but lacks the 5'-flanking promoter region and eight codons of the N-terminal coding region. The open reading frames of the complete gene and its corresponding cDNA clone are identical in sequence, indicating that this osmotin gene is indeed expressed in cotton. The cDNA insert is nearly full-length, lacking codons for four N-terminal amino acids. The two presumptive cotton osmotin preproteins can definitely be identified as PR5 proteins from their similarities to the deduced amino acid sequences of osmotin-like PR5 preproteins. The presumptive preproteins have N-terminal signal sequences, and the mature forms of the proteins would likely be targeted for extracellular secretion as neutral isoforms. In addition to basal TATA and CAAT promoter elements, other prospective promoter/enhancer elements, such as two ethylene response elements with the canonical sequence AGCCGCC, implicated as being positive regulatory elements in the expression of a number of PR proteins, occur in the 5’-flanking sequence of the full-length gene.