For regulatory issues and research purposes it would be desirable to have the ability to segregate transgenes in co-transformed maize. We have developed a highly efficient system to segregate transgenes in maize that was co-transformed using an Agrobacterium tumefaciens 2 T-DNA binary system. Three vector treatments were compared in this study; (1) a 2 T-DNA vector, where the selectable marker gene bar (confers resistance to bialaphos) and the beta-glucuronidase (GUS) reporter gene are on two separate T-DNA's contained on a single binary vector; (2) a mixed strain treatment, where bar and GUS are contained on single T-DNA vectors in two separate Agrobacterium strains; (3) and a single T-DNA binary vector containing both bar and GUS as control treatment. Bialaphos resistant calli were generated from 52 to 59% of inoculated immature embryos depending on treatment. A total of 93.4% of the bialaphos selected calli from the 2 T-DNA vector treatment exhibited GUS activity compared to 11.7% for the mixed strain treatment and 98.2% for the cis control vector treatment. For the 2 T-DNA vector treatment, 86.7% of the bialaphos resistant/GUS active calli produced R0 plants exhibiting both transgenic phenotypes compared to 10% for the mixed strain treatment and 99% for the single T-DNA control vector treatment. A total of 87 Liberty herbicide (contains bialaphos as the active ingredient) resistant/GUS active R0 events from the 2 T-DNA binary vector treatment were evaluated for phenotypic segregation of these traits in the R1 generation. Of these R0 events, 71.4% exhibited segregation of Liberty resistance and GUS activity in the R1 generation. A total of 64.4% of the R0 2 T-DNA vector events produced Liberty sensitive/GUS active (indicating selectable-marker-free) R1 progeny. A high frequency of phenotypic segregation was also observed using the mixed strain approach, but a low frequency of calli producing R0 plants displaying both transgenic phenotypes makes this method less efficient. Molecular analyses were then used to confirm that the observed segregation of R1 phenotypes were highly correlated to genetic segregation of the bar and GUS genes. A high efficiency system to segregate transgenes in co-transformed maize plants has now been demonstrated.