Two isolates of Pseudomonas fluorescens (2–79 and 13–79) from the USA were evaluated in the UK as biological control agents against Gaeumannomyces graminis var. tritici, the cause of take‐all in wheat. Biological control agents were applied as seed coatings in carboxymethyl cellulose (CMC) to seven wheat trials sown in 1987 and 1988 on fen peat and clay soils, and as peat‐based and microgranule formulations in one of these trials. In a trial of spring wheat on fen peat, all treatments with biological control agents reduced the percentage take‐all infection of crown roots and seminal roots, but the effects of only one isolate were statistically significant (P<0·05). Effects of biological control agents on infection rates in five other trials were not significant. In the trial in which application methods were compared, peat‐based inoculum initially appeared most effective but none of the treatments reduced take‐all significantly throughout the season. Application of biological control agents was associated with yield increases in several trials; these were not consistently associated with effects on take‐all. These results suggest that the isolates of P. fluorescens have potential to reduce take‐all and increase yields of wheat in the UK, but the beneficial effects are inconsistent. There is a need to develop isolates which reliably control severe take‐all in a variety of soil types.