The extensive impact cratering record on Mars combined with evidence from SNC meteorites suggests that a significant fraction of the surface is composed of materials subjected to variable shock pressures. Pressure-induced structural changes in minerals during high-pressure shock events alter their thermal infrared spectral emission features, particularly for feldspars, in a predictable fashion. To understand the degree to which the distribution and magnitude of shock effects influence martian surface mineralogy, we used standard spectral mineral libraries supplemented by laboratory spectra of experimentally shocked bytownite feldspar to deconvolve Thermal Emission Spectrometer (TES) data from six relatively large (>50 km) impact craters on Mars. We used both TES orbital data and TES mosaics (emission phase function sequences) to study local and regional areas near the craters, and compared the differences between models using single TES detector data and 3×2detector-averaged data. Inclusion of shocked feldspar spectra in the deconvolution models consistently improved the rms errors compared to models in which the spectra were not used, and resulted in modeled shocked feldspar abundances of >15% in some regions. However, the magnitudes of model rms error improvements were within the noise equivalent rms errors for the TES instrument [Hamilton V., personal communication]. This suggests that while shocked feldspars may be a component of the regions studied, their presence cannot be conclusively demonstrated in the TES data analyzed here. If the distributions of shocked feldspars suggested by the models are real, the lack of spatial correlation to crater materials may reflect extensive aeolian mixing of martian regolith materials composed of variably shocked impact ejecta from both local and distant sources.