Genetic engineering of cotton, thus far, has used Agrobacterium tumefaciens as the vehicle to deliver the foreign genes to the plant cell. The efficiency of transformation of the organism is dependent upon two factors, the binding of the bacterium to the cotton cell and the virulence of the Ti plasmid of the bacterium. In these experiments we tested some of the variability among cotton genotypes and bacterial strains for binding. Twelve strains of Agrobacterium representing each of the three biotypes were examined for their binding ability to eight genotypes of cotton suspension cells. Bacteria and plant cells were mixed together, rotated for three hours, and poured through a 10 um filter. Plant cells were washed with ddH₂O to remove unbound bacteria and then agitated, serially diluted, and plated. Colony counts were used to compare the number of colony forming units for each strain against each genotype. Binding parameter tests showed maximum binding at a pH of approximately 6.0. The binding of bacteria to the cotton cells was temperature-dependent, with maximum binding occurring around 26-28 C. There was no statistical evidence for interaction of cotton genotypes and bacteria strains. However, a difference at the 0.01 level was observed in the ratio of the number of bacteria in solution to the number of colony forming units for the different genotypes and bacteria strains. Differences were found at the 0.01 level between bacterial strains in the number of colony forming units, with A208 binding most efficiently. Delcot 344 was the cotton genotype associated with the greatest number of colony-forming units. Statistically significant differences were also observed at the 0.05 level between the cotton genotypes in their susceptibility to binding by the bacteria.