A series of Arctic sunphotometry ! ights were analyzed in terms of their multi-altitude, sub-micron (" ne mode) information content. A spectral deconvolution algorithm (SDA) and a " ne mode curvature algo- rithm (FMC) were applied to extract " ne mode and coarse mode optical depth as well as the effective radius of the " ne mode as a function of altitude. The " ne mode optical depth was differentiated as a function of altitude to retrieve vertical pro" les of " ne mode extinction coef" cient. These optical results were compared with volumetric altitude pro" les of " ne mode particle size distribution acquired by a UHSAS (Ultra High Sensitivity Aerosol Spectrometer). The results showed that layer-averaged extinction cross sections derived from the ratio of " ne mode optical depth to integrated UHSAS number density were signi" cantly larger than extinction cross sections derived from the application of Mie theory to the UHSAS particle size distributions. Vertical pro" les of extinction coef" cients derived from altitude derivatives of the " ne mode optical depth pro" les showed some correlation with the UHSAS extinction coef" cients. Pro" les of the " ne mode effective radius retrieved from the sunphotometry data were generally of similar magnitude to columnar averages of the UHSAS radii but no signi" cant trend with altitude could be detected. An example was given of a high altitude smoke plume whose presence eliminated any hope of correlating the volumetric sampling information with the sunphotometer pro" les. This is simply a statement of the obvious; that the atmospheric state must be stable in order to make such comparisons.