In an earlier study, treatment of radish seed with the bacterium Pseudomonas fluorescens WCS374 suppressed fusarium wilt of radish (Fusarium oxysporum f. sp. raphani) in a commercial greenhouse [Leeman et al., 1991b, 1995a]. In this greenhouse, the areas with fusarium wilt were localized or expanded very slowly, possibly due to disease suppressiveness of the soil. To study this phenomenon, fungi were isolated from radish roots collected from the greenhouse soil. Roots grown from seed treated with WCS374 were more abundantly colonized by fungi than were roots from nonbacterized plants. Among these were several species known for their antagonistic potential. Three of these fungi, Acremonium rutilum, Fusarium oxysporum and Verticillium lecanii, were evaluated further and found to suppress fusarium wilt of radish in a pot bioassay. In an induced resistance bioassay on rockwool, F. oxysporum and V. lecanii suppressed the disease by the apparent induction of systemic disease resistance. In pot bioassay s with the Pseudomonas spp. strains, the pseudobactin-minus mutant 358PSB- did not suppress fusarium wilt, whereas its wild type strain (WCS358) suppressed disease presumably by siderophore-mediated competition for iron. The wild type strains of WCS374 and WCS417, as well as their pseudobactin-minus mutants 374PSB- and 417PSB- suppressed fusarium wilt. The latter is best explained by the fact that these strains are able to induce systemic resistance in radish, which operates as an additional mode of action. Co-inoculation in pot bioassays, of A. rutilum, F. oxysporum or V. lecanii with the Pseudomonas spp. WCS358, WCS374 or WCS417, or their pseudobactin-minus mutants, significantly suppressed disease (except for A rutilum/417PSB- and all combinations with 358PSB-), compared with the control treatment, if the microorganisms were applied in inoculum densities which were ineffective in suppressing disease as separate inocula. If one or both of the microorganism(s) of each combination were applied as separate inocula in a density which suppressed disease, no additional suppression of disease was observed by the combination. The advantage of the co-inoculation is that combined populations significantly suppressed disease even when their individual population density was too low to do so. This may provide more consistent biological control. The co-inoculation effect obtained in the pot bioassays suggests that co-operation of P. fluorescens WCS374 and indigenous antagonists could have been involved in the suppression of fusarium wilt of radish in the commercial greenhouse trials.