Glucose transporters (GLUTs) are continuously recycled in 3T3-L1 cells and so insulin, through its action on phosphatidylinositol 3-kinase (PI 3-kinase), could potentially alter the distribution of these transporters by enhancing retention in the plasma membrane or acting intracellularly to increase exocytosis, either by stimulating a budding or a docking and fusion process. To examine the site of involvement of PI 3-kinase in the glucose transporter recycling pathway, we have determined the kinetics of recycling under conditions in which the PI 3-kinase activity is inhibited by wortmannin. Wortmannin addition to fully insulin-stimulated cells induces a net reduction of glucose transport activity with a time course that is consistent with a major effect on the return of internalized transporters to the plasma membrane. The exocytosis of GLUT1 and GLUT4 is reduced to very low levels in wortmannin-treated cells (approximately 0.009 min-1), but the endocytosis of these isoforms is not markedly perturbed and the rate constants are approx. 10-fold higher than for exocytosis (0.099 and 0.165 min-1, respectively). The slow reduction in basal activity following treatment with wortmannin is consistent with a wortmannin effect on constitutive recycling as well as insulin-regulated exocytosis. PI 3-kinase activity that is precipitated by anti-phosphotyrosine, anti(-)[insulin receptor substrate 1 (IRS1)] and anti-alpha-p85 antibodies show the same level of insulin-stimulated activity, approximately 0.5 pmol/20 min per dish of 3T3-L1 cells. Since the activities precipitated by all three antibodies are similar, it seems unlikely that a second insulin receptor substrate, IRS2, contributes significantly to the insulin signalling observed in 3T3-L1 cells. To examine whether insulin targets PI 3-kinase to intracellular membranes we have carried out subcellular fractionation studies. These suggest that nearly all the insulin-stimulated PI 3-kinase activity is located on intracellular, low-density, membranes. In addition, the association of PI 3-kinase with IRS1 appears to partially deplete the cytoplasm of alpha-p85-precipitatable activity, suggesting that IRS1 may redistribute PI 3-kinase from the cytoplasm to the low-density microsome membranes. Taken together, the trafficking kinetic and PI 3-kinase distribution studies suggest an intracellular membrane site of action of the enzyme in enhancing glucose transporter exocytosis.