The 2,4-diacetylphloroglucinol (Phi) biosynthesis locus, originally cloned from Pseudomonas flunrescens Q2-87, was introduced into the chromosome of seven fluorescent Pseudomonas strains (natural Phi producers: Q2-87R, Q65c-80R; Phi nonproducers: 2-79RNlO, 30-84R, M4-80R, Q69c-80R· and Q29z-80R) each with natural biocontrol aCtiVity against take-all of wheat or Pythium damping-off of chickpea. These transgenic pseudomonads were constructed by conjugating recipient pseudomonads with Escherichia coli SI7-1()..pir)(pUT::Km::phl) harboring a plasmid containing the PhI locus cloned within the mInitransposon TnS. Southern blot analysis revealed that the Phi locus was inserted randomly and uniquely into the recipient chromosome. HPLC analysis revealed that most transgenic pseudomonads were able to produce PhI, and their capability of Phi production varied. Transgenic isolates of both Q2-87R and Q65c-80R produced more Phi than their respective parental strains. Transgenic isolates of 2- 79RN 10, a natural producer of phenazine-l-carboxylic acid (PCA) , produced both PCA and Phi, whereas those of 30-84R, another phenazines producer, produced Phi but not phenazines any longer. The transgenic isolates of the Phi nonproducers M4-80R, Q29z-80R il.nd· Q69c-80R began to produce Phi after introduction of the Phi locus. Most transgenic pseudomonads exhibited consistent colony morphology and viability. However, the transgenic isolil.tes of Q69c-80R, a biocontrol strain without any apparent antibiotic production, were not stable in. those traits. They generally produced smaller and extraordinary colonies compared to the parent strain, indicating a high Phi production. When tested for the biocontrol activity with the transgenic isolates of Q69c-80R, they did not significantly improve the activity against take-all but required much smaller populations to obtain the significant suppression compared to the nontreated control.