In the article, the Monte Carlo method (MCM) has been characterized from the point of view of road accident reconstruction. This method lies in making repeated calculations with the use of the same deterministic mathematical model, but with picking out the values of specific parameters on a pseudo-random basis from within predefined ranges of uncertainty. The calculation results have been presented in the form of a probability density function similar, in terms of its graphical representation, to a bell-shaped curve; such a form facilitates the statistical interpretation of data and the uncertainty analysis. In particular, it is possible to narrow the range of results by rejecting the extreme areas of low probability. Examples have been presented, focused on the issues concerning the calculation of pre-impact velocities, location of the collision point on the road, and kinematic analysis (referred to as “time-distance analysis”) of the pre-impact phase of a pedestrian accident. In the collision analysis, both the reconstruction methods (based on the momentum conservation principle and on Marquard models of calculating the post-impact velocities) and simulation techniques (simulation of the impact and the dynamics of motion in the PC-Crash program) were employed. It has been shown that the area of the largest concentration of the Monte Carlo simulation results is actually the area of most common responses of the deterministic model used for the data ranges adopted, but not necessarily a reflection of the truth. The crucial point is to develop an adequate mathematical model of the physical phenomenon.