Purpose: The aim of this study was to determine a reliable compression testing method for cancellous jawbone specimens and to validate it by high-resolution finite element (FE) modeling based on micro-computerized tomography (CT) images of the specimens. Materials and Methods: Three series of human femoral bone samples were tested to establish a compression protocol for human jawbone cores. A CT scan of each bone sample was obtained. A simple destructive compression test was performed on the first series of 12 femoral bone samples (13 mm height and 6.1 mm diameter). The 5 femoral bone samples of the second series (13 mm height and 6.1 mm diameter) were constrained using end caps and subjected to 10 to 15 conditioning cycles before the destructive test from which the Youngs modulus (Emeas) was determined. The third series of 5 smaller femoral samples (8 mm height and 5.5 mm diameter) and the series of 5 jaw bone samples (8 mm height and 5.7 mm diameter) underwent the same testing protocol. FE models were created based on the CT images, and the simulated E-modulus (Ecalc) was calculated. Results: The intraclass correlation between Emeas and Ecalc corresponded to 0.74 for the first series of femoral bone samples, 0.96 for the second series, and 0.51 for the third series. For the jawbone samples, the intraclass correlation coefficient equaled 0.88. Conclusion: Reliable results for compression testing of cancellous jawbone can be obtained with cylindric specimens with a diameter of 5.7 mm, a length:diameter ratio 1.4, and flat top and bottom surfaces. The recommended compression method is constrained compression with 10 to 15 conditioning cycles, followed by a destructive test. Keywords: cancellous bone, compression, finite element modeling, jaws, micro-computerized tomography