Green fluorescent proteins (GFPs) are frequently used as marker and reporter systems to assess the fate and activity of microbial strains with the ability to degrade xenobiotic compounds. To evaluate the potential of this tool for tracking herbicide-degrading microorganisms in the environment a promoterless gfp was linked to the tfd C promoter, which is activated during degradation of the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D), and integrated into the chromosome of the 2,4-D-degrading strain Ralstonia eutropha JMP 134. The effects of the inserted gfp gene on the kinetics of 2,4-D degradation by R. eutropha in batch and chemostat culture were compared to those of the wild-type strain. In batch culture with 2,4-D as the only carbon and energy source the maximum specific growth rate of the gfp-marked strain did not differ significantly from the wild type. However, compared to the wild type, the 2,4-D steady-state concentration in 2,4-D-limited chemostat cultures of the gfp-marked strain was higher at all dilution rates tested. The reduced competitiveness of the gfp-marked strain at low substrate concentrations was confirmed in a competition experiment for 2,4-D in continuous culture at a dilution rate of 0.075 h-1. Reproducibly, the gfp-marked strain was displaced by the wild-type strain. The study clearly demonstrates that fitness of constructs cannot be assessed by measuring micro max with selected substrates in batch cultures only but that a thorough kinetic analysis is needed, which also considers slow, carbon-limited growth conditions as they occur in the environment.