Forensic pathologists who autopsy stab wound victims may be called upon to determine whether a specific weapon recovered as evidence caused the decedent's wounds. Some forensic training programs instruct pathologists to excise wounds and place them in formalin. There is little published information about the reliability of these assessments. To address these deficiencies, a porcine model was used to establish the limits of in situ knife wound measurements of wounds caused by 3 knives of different widths stabbed at 90 degrees to the skin surface. The results indicate that within the 95% confidence interval, most stab wounds when measured on the surface of the skin will be within 1 to 3 mm of the size of the knife blade. Four wounds from different anatomical areas of the pig were excised from the body, and after excision, their measurements differed from their in situ length. After fixation in a 10% buffered formalin solution, one wound stayed the same, one wound lengthened, one shrunk minimally, and the one from the thinner abdominal skin shrunk by 6 mm (11%). This study presents a porcine model that can be used to define parameters for testimony.