The future use of genetically modified microorganisms in the environment will be dependent on the ability to assess potential or theoretical risks associated with their introduction into natural ecosystems. To assess potential risks, several ecological parameters must be examined, including the impact of the introduced genetically modified organism on the microbial communities associated with the environment into which the introduction will occur. A 2-year field study was established to examine whether the indigenous bacterial communities of the rhizosphere and endorhiza (internal root tissues) were affected differently by the introduction of an unaltered wild type and its genetically modified derivative. Treatments consisted of the wild-type strain Pseudomonas fluorescens 89B-27 and a bioluminescent derivative GEM-8 (89B-27::Tn4431), Cucumber root or seed samples were taken 0, 7, 14, 21, 35, and 70 days after planting (DAP) in 1994 and 0, 7, 14, 28, 42, and 70 DAP in 1995. Samples were processed to examine the bacterial communities of both the rhizosphere and endorhiza. Over 7200 bacterial colonies were isolated from the rhizosphere and endorhiza and identified using the Sherlock System (Microbial ID, Inc.) for fatty acid methyl ester analysis. Community structure at the genus level was assessed using genera richness and Hill's diversity numbers, N1 and N2. The aerobic-heterotrophic bacterial community structure at the genus level did not significantly vary between treatments but did differ temporally. The data indicate that the introduction of the genetically modified derivative of 89B-27 did not pose a greater environmental risk than its unaltered wild type with respect to aerobic - heterotrophic bacterial community structure.