Sampling technique is a concern for standardization of thromboelastography (TEG), as it contributes towards preanalytical variation in TEG tracings. The objective of this study was to assess the influence of catheter lumen diameter on TEG parameters with kaolin-activated native blood in a swine haemorrhagic shock model, using standardized sampling techniques. We hypothesized that one or more TEG parameters will differ between venous samples collected from catheters with different diameter lumens. Ten male landrace swine were included in the study. Standard cut-down techniques on both femoral veins were performed and catheterized with a 14-gauge (14G) and a 20-gauge (20G) catheter of the same length. Blood withdrawal times were standardized to 3 s and TEG parameters (R, K, α, maximum amplitude, G and coagulation index) were measured or calculated. Significant differences were detected for R (14G < 20G; P = 0.002), K (14G < 20G; P = 0.026), maximum amplitude (14G > 20G; P = 0.02), G (14G > 20G; P = 0.001) and coagulation index (14G > 20G; P < 0.001). No significant differences were measured for α (P = 0.058). Bias values (method A: 20G; method B: 14G) for R, K, α, maximum amplitude, G and coagulation index were 1.00, 0.14, −1.88, −1.56, −1.39 and −1.07, respectively. Using TEG analysis, kaolin-activated native whole blood samples obtained from 14G catheters are hypercoagulable (clotted faster and stronger clot) compared with samples obtained from 20G catheters. Sampling methodology should be standardized and systematic when comparing TEG parameters.