The physical classification of crushed stone and gravel used in railway construction is based on their strength and endurance and is performed by a laboratory test method using a rotating drum or a mortar method. The values of fracture resistance calculated using the Los Angeles method and abrasion calculated using the Micro-Deval method show a corresponding correlation and require further investigation. Purpose. The development of a new method for measuring rock material fracture that is consistent with widely used standards while also being more comparable to real-world railway operating conditions. Certainly, both standard tests are essential for ensuring product homogeneity during production, so the new recommended method is only a supplement. Methodology. The Proctor device was used to induce so-called shock loads from above, similar to railway loading conditions. Unlike the standard method, the andesite material was placed in a standard cylinder in these tests. The samples were pre-screened and sorted; the specified weight was approximately 1,300 g, and the specified sizes of the individual particles were 6.3, 8.0 and 11.2 mm. Only prewashed and dried materials of NZ (fine crushed stone) or KZ (special crushed stone) from four different quarries (Tállya, Szob, Nógrádkövesd, Recsk) with different rock physics characteristics were considered. The Proctor compactor machine was used because of its calculable labor (19.86 J/impact) and the crushing effect of the calculable impacts (64, 128, 256 and 1,028 blows). Even after loading different numbers of impacts, homogeneous samples from different quarries were sieved to measure the masses of fragments per fraction. Findings. The set of measurements made it possible to establish a series of fragmentation and degradation curves for each of the three repeated measurements based on the composition of the material and the number of blows, which showed the degradation of samples with different physical and mechanical properties of the rock material and particle sizes. With an increasing number of impacts, the amount of crushed material in the sample increased, but the distribution of crushed material did not decrease evenly and proportionally as the number of impacts increased. Parameters and indices were also computed to identify various correlations (i. e., FV, d < 22.4, d < 0.5, d < 0.063 mm, CU, M ratio, ratio). Some of them (e. g., FV) needed to be changed, but they were predefined due to the nature of the tests. Originality. While many standard and alternative railway track ballast fragmentation test methods and measurement tools are available, this paper proposes a new laboratory method and demonstrates the specific measurement and application effectiveness. Practical value. In addition to standard tests that are already widely used, the new method for measuring the fractional composition of railway ballast can help simulate real-world operating conditions of a railroad track in the laboratory. This method will improve the safety of railway operations.