We previously reported that low intensity endurance training in sedentary patients suffering from the metabolic syndrome improves blood rheology, mostly due to a decrease in plasma viscosity correlated with an increase in cardiorespiratory fitness. We investigated whether these findings can be extended to type-2 diabetics. 22 diabetics (11 women and 10 men, age: 52.00 ± 2.9 yr, BMI: 32.47 ± 1.17 kg/m(2)) were tested before and after 2 months. Eight of them were trained (2 to 3×45 min/wk) at the power intensity where lipid oxidation reaches a maximum (LIPOX max) and thirteen served as controls. Over this period the only significant hemorheological effect of training was a decrease in RBC aggregation "M" (-1.25 ± 0.357 p = 0.01) in the trained group. Subjects who lost weight exhibited a decrease in plasma viscosity (from 1.46 ± 0.013 to 1.38 ± 0.02 p < 0.01). Changes in waist circumference are associated with changes in hematocrit (r =-0.952 p = 0.01); plasma viscosity (r =-0.91; p = 0.03); RBC aggregation ("M" r = 0.940; p = 0.02). Subjects can also be divided into those who improved their aerobic capacity VO(2max) and those whose VO(2max) decreased or remained unchanged. An increase in VO(2max) is associated with a decrease in whole blood viscosity (r =-0.79 p = 0.06) explained by an improvement in RBC rigidity "Tk" (r =-0.963 p = 0.002). This study suggests that in Type 2 diabetic patients: (a) viscosity factors might be less responsive to training than in non diabetic individuals; (b) visceral fat loss is the main determinant of changes in hematocrit, plasma viscosity and RBC aggregation;