Modeling of atmospheric aerosol optical characteristics is one of the most actively developed directions of atmospheric optics. However, results of these investigations turn out to be contradictory and in some cases troublesome. In our opinion, the reason is a limited number and low reliability of the initial data used for modeling as well as a certain gap in various directions of investigations. Possible ways to overcome the above—indicated intricacies and contradictions are considered in the paper.