Sustained hyperglycemia induces insulin resistance, but the mechanism is still incompletely understood. Glucosamine (GlcN) has been extensively used to model the role of the hexosamine synthesis pathway (HSP) in glucose-induced insulin resistance. 3T3-L1 adipocytes were preincubated for 18 h in media +/- 0.6 nmol/l insulin containing either low glucose (5 mmol/l), low glucose plus GlcN (0.1-2.5 mmol/l), or high glucose (25 mmol/l). Basal and acute insulin-stimulated (100 nmol/l) glucose transport was measured after re-equilibration in serum and insulin-free media. Preincubation with high glucose or GlcN (1-2.5 mmol/l) inhibited basal and acute insulin-stimulated glucose transport only if insulin was present during preincubation. However, only preincubation with GlcN plus insulin inhibited insulin-stimulated GLUT4 translocation. GLUT4 and GLUT1 protein expression were not affected. GlcN (2.5 mmol/l) increased cellular UDP-N-acetylhexosamines (UDP-HexNAc) by 400 and 900% without or with insulin, respectively. High glucose plus insulin increased UDP-HexNAc by 30%. GlcN depleted UDP-hexoses, whereas high glucose plus insulin increased them. Preincubation with 0.5 mmol/l GlcN plus insulin maximally increased UDP-HexNAc without affecting insulin-stimulated or basal glucose transport. GlcN plus insulin (but not high glucose plus insulin) caused marked GlcN dose-dependent accumulation of GlcN-6-phosphate, which correlated with insulin resistance of glucose transport (r = 0.935). GlcN plus insulin (but not high glucose plus insulin) decreased ATP (10-30%) and UTP (>50%). GTP was not measured, but GDP increased. Neither high glucose plus insulin nor GlcN plus insulin prevented acute insulin stimulation (approximately 20-fold) of insulin receptor substrate 1-associated phosphatidylinositol (PI)-3 kinase. We have come to the following conclusions. 1) Chronic exposure to high glucose or GlcN in the presence of low insulin caused insulin resistance of glucose transport by different mechanisms. 2) GlcN inhibited GLUT4 translocation, whereas high glucose impaired GLUT4 "intrinsic activity" or membrane intercalation. 3) Both agents may act distally to PI-3 kinase. 4) GlcN has metabolic effects not shared by high glucose. GlcN may not model HSP appropriately, at least in 3T3-L1 adipocytes.