The proportion of vibrio E, dominant in healthy turbot larvae, was artificially increased in the rotifer enrichment medium. The maid effect of this vibrio enrichment was to improve the resistance of larval turbot challenged with a pathogenic strain of Vibrio splendidus, vibrio P. The improvement of the survival rates was significant with 1.5 x 10(6) CFU of vibrio E added per mi of enrichment medium, and it was still maximum at the hi hest dose tested (5 x 10(7) CFU.ml(-1); 80 +/- 2% of survivors after 48 h of challenge, versus 50 +/- 2 % for the control group). Further in vitro characterisation showed that vibrio E was able to grow in the presence of the chelating agents EDDHA and EDTA, while producing siderophore, as revealed by chrome azurol S dyeing. The pathogenic vibrio P did not grow on iron-depleted medium, and it did not seem to produce siderophore. Consequently, a second experiment attempted to appraise the role of iron availability in the interaction between the pathogenic and probiotic vibrios. To this end, turbot larvae were fed with rotifers treated by the bacterial siderophore deferoxamine. This treatment significantly increased the resistance of larvae challenged with vibrio P in comparison with the control group, but to a lesser extent than the group enriched with vibrio E. There was also an increase of the growth rate due to either the siderophore treatment or the vibrio enrichment, as well. No further improvement was observed when the rotifers were simultaneously both treated and enriched. It was concluded that the probiotic effect of vibrio E may be at least partly due to competition for iron with the pathogen.