The hypothesis that mycorrhizal colonization improves the soil-root conductance in plants was experimentally tested in a growth chamber using pot cultures of Agrostis stolonifera L. colonized by Glomus intraradices. Plants were grown in 50-l pots filled with autoclaved sand/silt soil (1:1), with and without the mycorrhizal fungus. Within the mycorrhizal treatment, half of the pots remained well watered, while the other half was subjected to a progressive water deficit. Soil water potential (estimated as plant water potential measured at the end of the dark period), xylem water potential measured at the tiller base, transpiration rate, and soil water content were monitored throughout the experiment. Soil-root hydraulic conductance was estimated as the ratio between the instantaneous transpiration rate and the soil and xylem water potential difference. To obtain cultures with similar nutritional status, the P in the modified Hoagland's nutrient solution was withheld from the inoculated pots and applied only once a month. Even though there were no differences on growth or nutrient status for the mycorrhizal treatments, water transport was enhanced by the inoculum presence. Transpiration rate was maintained at lower xylem water potential values in the presence of mycorrhizae. The analysis of the relationship between soil-root hydraulic resistance and soil water content showed that mycorrhizal colonization increased soil-root hydraulic conductance as the soil dried. For these growing conditions, this effect was ascribed to the range of 6-10%.