To investigate the influence of specimen geometry and gauge length on the tensile properties of SiC_f/SiC composites， specimens are cut from the same plate： dog-bone-shaped specimens （Type A） and Y-tabbed specimens with two gauge lengths （Types B and C）. Room-temperature tensile tests are performed. Digital image correlation （DIC） is used to analyze damage evolution， and fracture characteristics are examined to thoroughly investigate the effects of specimen configuration and gauge length on tensile test results. It is found that specimen geometry and gauge length have a minor influence on the linear segment of the tensile stress-strain curve. However， the geometry significantly affects the ultimate tensile strength and fracture strain. Type A specimens， employing flat clamping， introduce an additional bending moment during loading， resulting in significantly lower ultimate tensile strength and fracture strain compared with Types B and C specimens， which use Y-tabbed suspension gripping. The difference in tensile properties between specimens with gauge lengths of 25 mm （Type B） and 15 mm （Type C） is less than 9%. This confirms that the size effect can be neglected when the gauge length exceeds 10 times the characteristic length （L_c=0.49 mm）. The study demonstrates that Y-tabbed suspension specimens enable accurate determination of the material’s true properties， and a gauge length greater than 5 mm can eliminate the influence of the size effect for the present SiC_f/SiC composites. This research provides experimental evidence and theoretical support for the performance evaluation of aero-engine hot-section components and the design of specimens for in-situ sampling.