Arid soils pose unique problems during measurement and interpretation of particle-size distributions (PSDs) because they often contain high concentrations of water-soluble salts. This study investigates the effects of sulfate and carbonate minerals on grain-size analysis by comparing analyses in water, in which the minerals dissolve, and isopropanol (IPA), in which they do not. The presence of gypsum, in particular, substantially affects particle-size analysis once the concentration of gypsum in the sample exceeds the mineral’s solubility threshold. For smaller concentrations particle-size results are unaffected. This is because at concentrations above the solubility threshold fine particles cement together or bind to coarser particles or aggregates already present in the sample, or soluble mineral coatings enlarge grains. Formation of discrete crystallites exacerbates the problem. When soluble minerals are dissolved the original, insoluble grains will become partly or entirely liberated. Thus, removing soluble minerals will result in an increase in measured fine particles. Distortion of particle-size analysis is larger for sulfate minerals than for carbonate minerals because of the much higher solubility in water of the former. When possible, arid soils should be analyzed using a liquid in which the mineral grains do not dissolve, such as IPA, because the results will more accurately reflect the PSD under most arid soil field conditions. This is especially important when interpreting soil and environmental processes affected by particle size.