Machine guards provide protection against ejection of parts during operation, such as chips or workpiece fragments. They are considered safe if the impact resistance is at least as high as the resulting projectile energy in the worst case of damage. To protect the machine operator, the impact resistance of machine guards is determined according to ISO standards. The bisection method can be used to determine the impact resistance through impact tests. However, this method is inaccurate for a small number of impact tests and does not provide an indication of uncertainties in the determination. Moreover, the result of testing is validated in different ways depending from the standard utilized for testing.Relevant uncertainties affecting impact testing and a new probabilistic approach for assessing the impact resistance using the Recht & Ipson equation are presented. With multiple impact tests at different initial velocities a Recht & Ipson best-fit curve and a confidence interval for a ballistic limit can be obtained, which is used to determine the impact resistance by defining a velocity reduction coefficient. This method can be applied to any machine guard made of ductile material. This paper validates the Recht & Ipson method by performing impact tests with a standardized 2.5 kg projectile on polycarbonate sheets of different thicknesses. Determination of the ballistic limit showed good agreement with experimental results. With the ballistic limits, the velocity reduction coefficients have been found to determine the impact resistances. Therefore, an alternative method for standardized tests to determine the impact resistance was found.