Chronic hyperglycemia promotes the development of insulin resistance. The aim of this study was to investigate whether cellular insulin resistance is secondary to the diabetic state in human type 2 diabetes. Subcutaneous fat biopsies were taken from 3 age-, sex-, and body mass index (BMI)-matched groups with 10 subjects in each group: type 2 diabetes patients with either good (hemoglobin A1c [HbA1c] [lt ] 7%, G) or poor (HbA1c [gt ] 7.5%, P) metabolic control and healthy control subjects (C). Insulin action in vitro was studied by measurements of glucose uptake both directly after cell isolation and following a 24-hour incubation at a physiological glucose level (6 mmol/L). The relationship with insulin action in vivo was addressed by employing the euglycemic clamp technique. Freshly isolated fat cells from type 2 diabetes patients with poor metabolic control had [sim ]55% lower maximal insulin response (1,000 [mu ]U/mL) on glucose uptake (P [lt ] .05) compared to C. Cells from P were more insulin-resistant (P [lt ] .05) than cells from G at a low (5 [mu ]U/mL) but not at a high (1,000 [mu ]U/mL) insulin concentration, suggesting insulin insensitivity. However, following 24 hours of incubation at physiological glucose levels, insulin resistance was completely reversed in the diabetes cells and no differences in insulin-stimulated glucose uptake were found among the 3 groups. Insulin sensitivity in vivo assessed with hyperinsulinemic, euglycemic clamp (M-value) was significantly associated with insulin action on glucose uptake in fresh adipocytes in vitro (r = 0.50, P [lt ] .01). Fasting blood glucose at the time of biopsy and HbA1c, but not serum insulin, were negatively correlated to insulin's effect to stimulate glucose uptake in vitro (r = [minus ]0.36, P = .064 and r = [minus ] 0.41, P [lt ].05, respectively) in all groups taken together. In the in vivo situation, fasting blood glucose, HbA1c, and serum insulin were all negatively correlated to insulin sensitivity (M-value; r = [minus ]0.62, P[lt ] .001, r= [minus ]0.61, P[lt ] .001, and r = [minus ]0.56, p [lt ] .01, respectively). Cell size, waist-to-hip ration (WHR), and BMI correlated negatively with insulin's effect to stimulate glucose uptake both in vitro (r = [minus ]0.55, P [lt ] .01, r = [minus ]0.54, P [lt ] .01, and r = [minus ]0.43, P [lt ] .05, respectively) and in vivo (r = [minus ]0.43, P [lt ] .05, r = [minus ]0.50, P [lt ] .01, and r = [minus ]0.36, P [lt ] .05, respectively). Multiple regression analyses revealed that adipocyte cell size and WHR independently predicted insulin resistance in vitro. Furthermore, insulin sensitivity in vivo could be predicted by fasting blood glucose and serum insulin levels. We conclude that insulin resistance in fat cells from type 2 diabetes patients is fully reversible following incubation at physiological glucose concentrations. Thus, cellular insulin resistance may be mainly secondary to the hyperglycemic state in vivo.