Objective To investigate the effect from different pore sizes of co culture inserts on the permeability of biomacromolecules through polyethylene terephthalate (PET) membrane so as to solve the key technology problem in mechanobiology experiment on vascular cells. Methods Inserts with 0.4 μm and 1.0 μm pores on the PET membrane were studied using flow chamber system. Low shear stress was subjected to the co-cultured system of endothelial cell (EC)/vascular smooth muscle cell (VSMC) and the concentration of platelet-derived growth factor BB (PDGF-BB) was detected by ELISA. Under the static condition, vascular cells were cultured on the plate (with no cell on PET membrane), on the outer side of PET membrane, and on the both sides of PET membrane, respectively. Then the recombinants PDGF-BB (rPDGF-BB) were added on the different sides of PET membrane. Western blotting was used to detect the change in expressions of p-ERK1/2, p-Akt and Lamin after cells were stimulated by rPGDF BB. Results After low shear stress subjection for 12 h, the concentration of PDGF-BB in the medium from VSMC side was significantly higher than that from EC-side. rPDGF-BB passed through 0.4 μm and 1.0 μm pores on the PET membrane and modulated expressions of p-ERK1/2, p-Akt and Lamin A in cells cultured on the opposite side of PET membrane and cells cultured on the plate separately. When cells were cultured on the both sides of PET membrane, rPDGF-BB only stimulated cells cultured on the same side of 0.4 μm pores on PET membrane, but had no specific effect on cells cultured on the opposite side. Conclusions PET membrane with both 0.4 μm and 1.0 μm pores was permeable to PDGF-BB, and cells cultured on the membrane could affect the permeability. The efficiency of PDGF BB passing through 0.4 μm pores was significantly repressed with cells cultured on the both sides, which was more similar to that in vivo.