Kolsky Bar systems are subjected to inherent system error as all measurement devices are. This is especially true in that as the bar diameter decreases, the system becomes more sensitive to errors such as friction and misalignment. In this work we present a technique for identifying and quantifying the error of a Kolsky system. We also present a method of generating statistically significant bounds for Kolsky systems so that anomalous or improperly executed experiments can be quantitatively identified. This method does not rely on the intuition of the experimentalist to identify an anomalous experiment. After presenting our method for error identification, a series of tests are performed on 2024Aluminum alloy samples. A method is then presented where the system error, as well as some error contributed by a variance in sample dimension, are removed from the calculated error related to the stress on the samples. The result shows the effective variance of the sample response is quite high in the elastic loading period, but reduces when plasticity dominates. This is attributed to the presence of high frequency content in the travelling elastic waves which cannot be accurately measured currently, but is effectively damped out when plastic deformation dominates.