Using spectral images recorded by the OMEGA instrument on Mars Express (Observatoire pour la Minéralogie, l'Eau, les Glaces et l'Activité), we are able to derive physical properties of aerosols in water-ice clouds on Mars for a distribution of pixels over an observed cloud formation. These properties, mean effective radius, reff, and optical depth (at 0.67 μm), τi, were used to estimate the water ice-column (WIC), and we found an empirical relationship between the WIC and an ice cloud index (ICI). The overall mean of retrieved reff is ∼2.2 μm, with a standard deviation of 0.8 μm, and cloud formations with reff between 4.4 and 5.4 μm are observed. The optical depth varies between 0.2 and 2.0. The OMEGA spectra are primarily sensitive to water ice mass due to absorption, and we find that the ICI, very easy to compute, is a good proxy for the mass of the water-ice column (WIC) along the optical line of sight. Our retrieval of physical properties is limited in time (to before 2010) by the exhaustion of coolant for one of the OMEGA channels, and in space (to equatorial observations between 140∘W and 90∘E) by the availability of surface albedo measurements. However, we used the ICI to compute WIC values for the entire OMEGA data set, which has near-global coverage for Mars years 26–32, and we present a climatology of the WIC derived from the OMEGA data, which features enhancements on the order of 1.2–1.6 pr. μm over the aphelion cloud belt, and 1.5–2.5 pr. μm over the polar hoods. The data set analyzed is for observations between 140°W and 90°E, and between 35∘S and 35∘N. No restriction is placed on season, but the majority of cloudy observations were during the aphelion period from Ls 35∘ to 135∘. This work was motivated by the ability of the OMEGA instrument to observe the distribution of water-ice cloud physical properties, and by the availability of new a priori data sets, especially multi-spectral, aerosol-free surface albedo retrieved from a subset of the OMEGA data featuring a cloud-free sky. The main limitations of the retrieval algorithm are linked to the uncertainties on surface albedo, the dust opacity, and the quantity of water-ice suspended in the atmosphere, which can lead to spectral fits with lower accuracy or unrealistic results. We present distributions of each retrieved parameter, goodness of fit, ICI, and cloud mass, and our investigation of relationships between each parameter. Our approach was to maximize the amount of data analyzed, apply stringent data quality cuts and take a statistical approach to interpretation.