X-ray halos around point sources provide information regarding interstellar grains. The scattering is significantly affected by (a) the energy of the radiation; (b) the optical depth of the scattering, because of the effects of multiple scattering; (c) whether or not the grains are porous or fluffy; and (d) the spatial distribution of dust along the line of sight. The scattering is not strongly affected by the amount of very small grains (or large molecules) present, by the differences among the various grain models which employ compact (nonporous) grains, or by the chemical composition of the grains if there are no absorption edges resolved by the observations. Multiple scattering for optical depths tau-sca almost-equal-to 1 strongly affects the outer wings, dominating the total flux for tau-sca > 1.3. Diagnosing grain models from optically thick scattering requires good energy resolution for two reasons: (1) the importance of multiple scattering decreases rapidly with energy, making the halo narrower, and (2) the single scattering halo narrows with energy. The distribution of dust along the line of sight affects the inner part of the halo, making the halo more peaked toward the central point source if the dust is concentrated near the source. The halo is also affected by the distribution of dust along the line of sight, especially if the source lies within a zone with appreciably lower than average dust density, such as well out of the Galactic plane. A plot of log (R12) = log [I(sca)(10")/I(sca)(30")] versus log (R23) = log [I(sca)(30")/I(sca)(120")] helps to separate the effects of such a distribution from the effects of multiple scattering and varying grain-size distributions. The spatial distribution primarily affects R12; the scattering and size distribution affect R23 strongly. Since the optical depth of the scattering can be inferred from the energy dependence of the halo strength, the effects of multiple scattering can be determined separately from grain-size distribution. The observations of an occultation of an X-ray source near the Galactic center by the Moon can be explained only if the multiple scattering at E < 2 keV is quite important. Models of diffuse dust [R(V) = A(V)/(B-V) almost-equal-to 3] can explain the published one-parameter fit to the results. A porous-grain dense-cloud dust model (R(V) = 4.8) predicts too fast a decrease in scattered intensity with angle. Presumably, large grains are actually less porous than the 80% of vacuum assumed in the model.