Abstract Two laboratory experiments were used to investigate the effect of the earthworm Lumbricus terrestris on transport of genetically marked Pseudomonas fluorescens inocula through soil microcosms. The microcosms comprised cylindrical cores of repacked soil with or without earthworms. Late log-phase cells of P. fluorescens, chromosomally marked with lux genes encoding bioluminescence, were applied to the surface of soil cores as inoculated filter paper discs. In one experiment, 5 and 10 days after inoculation, cores were destructively harvested to determine concentrations of marked pseudomonads with depth relative to the initial inoculum applied. Transport of the bacteria occurred only in the presence of earthworms. In a second experiment cores were subjected to simulated rainfall events 18h after inoculation with lux-marked bacteria at 3-day intervals over a 24-day period. Resulting leachates were analysed for the appearance of the marked bacteria, and after 28 days cores were destructively harvested. Although some marked cells (less than 0.1% of the inoculum applied) were leached through soil in percolating water, particularly in the presence of earthworms, the most important effect of earthworms on cell transport was through burial of inoculated litter rather than an increase in bypass flow due to earthworm channels.