Although D-glucose increased the root growth of Arabidopsis seedlings, D-allose (a d-glucose epimer at the third carbon atom) inhibited the root growth at concentrations >0.1 $mmol;L^{-1}$ and the inhibition increased with increasing D-allose concentrations. Allitol (a reduction product of D-allose) did not show any significant effect on the growth. The addition of D-glucose into the growth medium of Arabidopsis reversed the D-allose-induced growth inhibition, which suggests that the inhibition is not caused by the toxicity of the accumulation of D-allose and/or its metabolites in the seedlings. D-Allose is phosphorylated by hexokinase, using ATP and phosphate, to allose-6-phosphate, with no known capacity for further metabolism. The addition of phosphate into the growth medium did not affect the D-allose-induced growth inhibition and D-allose did not reduce the ATP level in the roots. These results suggest that the inhibition is not due to phosphate starvation and ATP depletion. D-Mannoheptulose, a specific competitive inhibitor of hexokinase, defeated the D-allose-induced growth inhibition. Hexokinase is known to have a sugar-sensing function and possibly triggers a signal cascade, resulting in the change of several gene expressions. Therefore, the phosphorylation of D-allose by hexokinase might trigger a signal cascade, resulting in the inhibition of Arabidopsis root growth. This is probably a useful model system for studies of the hexokinase-mediated sugar-sensing function and for developing new types of weed-control agents.