The fiber type-specific expression of skeletal muscle GLUT4 and the effect of 2 weeks of low-intensity training were investigated in 8 young untrained male subjects. Single muscle fibers were dissected from a vastus lateralis biopsy sample. Based on myosin heavy chain (MHC) expression, fibers were pooled into 3 groups (MHC I, MHC IIA, and MHC IIX), and the GLUT4 content of 15-40 pooled fibers was determined using SDS-PAGE and immunological detection. The GLUT4 content in pooled muscle fibers expressing MHC I was ~20% higher (P < 0.05) than that in muscle fibers expressing MHC IIA or MHC IIX. No difference in GLUT4 could be detected between fibers expressing MHC IIA or MHC IIX. Two weeks of exercise training increased (P < 0.05) the peak power output of the knee extensors by 13%, the maximal activities of citrate synthase and 3-hydroxyacyl-CoA dehydrogenase by 21 and 18%, respectively, and the GLUT4 protein content by 26% in a muscle homogenate. Furthermore, a 23% increase (P < 0.05) in GLUT4 was seen in fibers expressing the MHC I isoform after exercise training for 2 weeks. No change was seen in fibers expressing MHC IIA or MHC IIX. In conclusion, our data directly demonstrate that GLUT4 is expressed in a fiber type-specific manner in human skeletal muscle, although fiber type differences are relatively small. In addition, low-intensity exercise training recruiting primarily fibers expressing MHC I increased GLUT4 content in these fibers but not in fibers expressing MHC IIA or MHC IIX, indicating that GLUT4 protein content is related more to activity level of the fiber than to its fiber type, which is defined by expression of contractile protein. Diabetes 49:1092-1095, 2000 I n both rats and humans, glucose transport has been shown to be the rate-limiting step in muscle glucose uptake under most conditions (1,2). Glucose transport in mammalian skeletal muscle is almost exclusively mediated by the insulin-and contraction-regulatable glucose transporter GLUT4 (3). In both rodents and humans, different muscles have been shown to exhibit large differences in their GLUT4 content (4-9), and in rats, this variation is often associated with differences in insulin-stimulated glucose uptake (4-6,8). Because different muscles are composed of a mixture of several different muscle fiber types in both rats and humans (10), it is possible that a relationship exists between fiber type composition and GLUT4 content. Such a relationship has been proposed in human muscle based on indirect evidence (11,12), but in reality, it is unknown at present if the GLUT4 content differs between fiber types in human skeletal muscle.It is also possible that the differences in GLUT4 content and insulin-stimulated glucose uptake are more related to activity level. Changes in the skeletal muscle activity level have been shown to be important regulators of the GLUT4 content in rats (13-16). In humans, athletes have more GLUT4 than untrained age-matched control subjects (11,17,18), and in both normal healthy control subjects and individuals with de...