The efficiency of a percussive process for fragmentation of rock and similar materials has been studied experimentally. The percussive system comprised a cylindrical hammer and a cylindrical bit with the same characteristic impedance. The bit was terminated with a wedge. In front of the wedge there was a heavy block of concrete. The length of the bit and the initial gap between bit and concrete were varied systematically. The force versus penetration relationship and the work of fragmentation were determined in each test using a new technique based on measurement of strains at two cross-sections of the bit. Each test was simulated individually using a previously developed one-dimensional model. The results of simulations and experimental tests were found to agree well.