SUMM.^RYOur objectives were (1) to determine if arbuscular mycorrhizal symbiosis could modify leaf response to nonhydraulic root-to-shoot communication of soil drying in Sorghum hicolor (L.) Moench, and (2) to compare the sensitivity of leaf growth and stomatal conductance (C^) to the non-hydraulic signal. SteJlings were grown in a greenhouse with root systems split among four pots. Treatments were applied in a 2 (colonized or not colonized by Glomus iittraradices Schenck & Smith) x 2 (roots isevered or dried) x 4 (roots in 0, 1, 2 or 3 pots dried or seyered) experimental design. Plants with toots in three pots dried or severed showed reduced leaf elongation, C^ and leaf water potential (4'') compared with fully watered (control) plants and thus were prohably hydraulically affected by root treatnient. Drying ot severing roots in one pot did not affect leaf elongation. C^ or H* in either mycorrhiza! or non-mycorrhizal plants. In non-mycorrhizal plants having two pots dried, final leaf area and toiiil ieaf length were reduced by 18 and 10",,, respecti\'eh', relative to controls. Stomatal conductance of these half-dried nonmycorrhizal plants remained unchanged, suggesting that the decrease in leaf growth was not hydrauiically induced. Non-mycorrhizal plants ha\'ing root.s severed in two pots contmued to have ieaf growth sirniiar to that of the controls, suggesting that growth reductions in half-dried non-mycorrhizal pJants did not result from a reduction in root water gathering capacity. Mycorrhizal symbiosis appeared to eliminate inhibition of leaf growth that uaH not hydraulically induced, because mean leaf area and total leaf length were not reduced in half-dried mycorrhizal plants, relative to controls. However, final leaf area of mycorrhizal plants having one or two pots dried was negatively correlated with the produtt of drying root mass and the time for which roots were exposed to mild drought, suggesting that mycorrhizal plants were also susceptible to growth inhibition that was not hydraulically induced. Reductions in leaf extension rate that were not hydraulically induced, when viewed as a function of actual .soil matric potential, were similar in mycorrhiza] and non-mycorrhizal plants, suggesting that the differences in overall growth inhibition between half-dried mycorrhizal and non-mycorrhizal plants may have been related to difFerences in soil drying rate.