Investigation of weapons noise exposure outdoors necessitates including the impact of ground reflections. When analyzing the reflections’ impact on A-weighted equivalent level (LAeq), the contribution of the reflected waveform to the total level can often be acceptably modeled using a Friedlander equation-based source model and an ideal reflection based on the measurement geometry. However, the Friedlander model does not adequately model the negative phase of a gunshot waveform, making it inaccurate for locations where the waveform of the direct blast and the waveform of the reflection begin to overlap temporally. This merging of the waveforms for locations where the direct and ground-reflected path length difference is relatively small also makes it difficult to separate the waveforms and calculate the reflected wave’s contribution to noise exposure. As such, this paper investigates different approaches to estimating the contribution of the ground reflected wave to the LAeq, using direct and reflected waveforms in regions where they are well separated and then adjusting timing and amplitude for different geometries where merging takes place. Results reveal regions where the ground reflection contributes negligibly or even negatively to the total LAeq, which is not predicted by the Friedlander model nor by simple energy-addition approaches. [Work supported by Air Force Research Laboratory through Ball Aerospace Technology Corporation.]