Abstract: To address the practical difficulties associated with field sampling and laboratory machining of standard cubic specimens containing joints, this study draws on the principles of directional drilling and borehole coring technology to develop curved loading platens for jointed core specimens. Laboratory experiments are conducted to verify the feasibility of using borehole cores for direct shear testing of joint shear strength. In combination with numerical simulations, the potential effects of core diameter, joint roughness, and related factors on the test results are quantitatively evaluated, and a calibration model for the shear strength parameters is subsequently established. The laboratory results indicate that, for planar joints, the discrepancy in test results between standard cubic specimens and core specimens with a diameter of 50 mm is less than 2%, suggesting that core specimens can directly provide shear strength parameters that satisfy code requirements. Numerical simulations further show that, for rough joints, certain differences exist in the mechanical response and failure patterns between cubic and core specimens, and these differences become more pronounced with decreasing core diameter or increasing joint roughness. Nevertheless, the shear strength obtained from core specimens generally conforms to the Mohr-Coulomb strength criterion, although the corresponding strength parameters exhibit a negative correlation with core diameter and a positive correlation with joint roughness. On this basis, through regression analysis and polynomial modeling, mathematical relationships are proposed between the calibration coefficients for cohesion and friction coefficient and the core diameter as well as joint roughness. These relationships enable the conversion of shear strength parameters measured from core specimens into the equivalent results for standard cubic specimens. Mesoscopic analysis reveals that the discrepancy in joint strength obtained from cubic and core specimens mainly arises from differences in the statistical distribution of the local friction coefficient along the direction perpendicular to shearing. A reduction in core diameter intensifies the central polarization characteristics, whereas an increase in joint roughness markedly amplifies the fluctuation range. The findings effectively validate the applicability of borehole cores for determining joint shear strength and provide a new perspective and practical solution for strength testing of engineering rock masses.