An experimental study was performed to determine the feasibility of optically distinguishing coal particles from water droplets for the purpose of reducing the mass concentration artifact from water sprays and steam reported for optical dust monitors. A single-particle multiangle optical detector was used to measure the scattering of mono chromatic, circularly polarized light by mixtures of nonspherical bituminous coal dust particles and 0.5 µm dioctyl sebacate (DOS) droplets, which were used for modeling small, spherical water drop lets. A total of 20 different mixtures with known coal number fractions ranging from 6 to 90 percent were measured by the detector using two data acquisition methods known as peak-detection and digital signal evaluation (DSE). Variability in azimuthal scattering by single particles, which is an indicator of the degree of sphericity, was used to discern the fraction of coal particles present in each aerosol mixture. Number concentrations of coal and DOS aerosols were measured with an optical particle counter to derive expected coal number fractions. Measured coal number fractions compare well with expected values with average overestimates of 0.040 for peak-detection 0.087 for DSE. A twofold increase in the average peak-detection overestimate resulted when overlap in DOS and coal azimuthal variability distributions was neglected in the inference of coal fractions, but no significant change in the DSE values was observed. Problems to be overcome in developing a single-particle optical dust monitor with shape distinction capability for use in mining and industrial environments are briefly discussed.