The successful release of wild type and genetically modified micro-organisms into soil has been impeded because efficacy is unpredictable. Evidence is presented showing that the plant-inoculum interaction can be strongly influenced by cropping system management or soil type. Rhizobacteria that inhibited the root growth of Poa annua and Lolium perenne were isolated from soil. The isolates, an inhibitory isolate (W7+) and a non-inhibitory mutant (G2Y), were identified as Pseudomonas putida. The rhizobacteria were used as soil inocula in a completely randomized block design that included Poa and Lolium, four soils, and three inoculum carriers. Of the possible 15 interactions, 10 significantly affected both the root and shoot growth. These results suggest that field trials conducted without consideration of soil quality factors and carrier delivery systems will be of dubious value when evaluating the performance of allochthonous micro-organisms in soil.