The increasing demand for bone grafts, combined with their limited availability and potential risks, has led to much new research in bone tissue engineering. Current strategies of bone tissue engineering commonly use cell-seeded scaffolds and flow perfusion bioreactors to stimulate the cells to produce bone tissue suitable for implantation into the patient's body. The aim of this study was to quantify and compare the wall shear stresses in two bone tissue engineering scaffold types (collagen-glycosaminoglycan (CG) and calcium phosphate) exposed to fluid flow in a perfusion bioreactor. Based on micro-computed tomography images, three-dimensional numerical computational fluid dynamics (CFD) models of the two scaffold types were developed to calculate the wall shear stresses within the scaffolds. For a given flow rate (normalized according to the cross-sectional area of the scaffolds), shear stress was 2.8 times as high in the CG as in the calcium-phosphate scaffold. This is due to the differences in scaffold geometry, particularly the pore size (CG pore size approximately 96 microm, calcium phosphate pore size approximately 350 microm). The numerically obtained results were compared with those from an analytical method that researchers use widely experimentalists to determine perfusion flow rates in bioreactors. Our CFD simulations revealed that the cells in both scaffold types were exposed to a wide range of wall shear stresses throughout the scaffolds and that the analytical method predicted shear stresses 12% to 21% greater than those predicted using the CFD method. This study demonstrated that the wall shear stresses in calcium phosphate scaffolds (745.2 mPa) are approximately 40 times as high as in CG scaffolds (19.4 mPa) when flow rates are applied that have been experimentally used to stimulate the release of prostaglandin E(2). These findings indicate the importance of using accurate computational models to estimate shear stress and determine experimental conditions in perfusion bioreactors for tissue engineering.