Two pools of hexokinase activities differing in sensitivity to ADP inhibition were characterised in maize roots. In order to evaluate how glucose utilisation could be affected by these hexokinases, glucose-6-P and NDP-5'-sugar levels were measured after a D-[U-14C]glucose pulse in root extracts in the presence of 0 or 1 mM ADP. Analysis of radio-labelled activated sugars by paper chromatography revealed that: (1) without ADP, nearly 20% of the 14C appeared in NDP-5'-sugars; (2) 0.1 mM ADP inhibited 14C-NDP-5'-sugar formation by 85%; and (3) with 1 mM ADP, 14C-NDP-5'-sugars were undetectable, but substantial (14%) 14C accumulated as glucose-6-P. Mannoheptulose, a hexokinase inhibitor, blocked the NDP-5'-sugar formation, but did not modify the amount of 14C-glucose-6-P in root extracts either with or without ADP. The analysis of the hexokinase activities with 0.8 mM glucose in maize root extracts showed that: (1) mitochondrial hexokinase activity was totally inhibited by 30 mM mannoheptulose; and (2) the cytosolic hexokinase was inhibited by only 30%. These data suggest that NDP-5'-sugar synthesis is sensitive to ADP fluctuations and that mannoheptulose affects preferentially the mitochondrial-bound hexokinase, but the cytosolic form is less sensitive. We propose that the mitochondrial hexokinase is the main energy charge sensor in this pathway in maize.