Although ice clouds play an important role in the energy budget balance of Earth-Atmosphere system, they are poorly represented in most of current general circulation models. The main difficulty with developing an ice cloud parameterization scheme for GCMs is that the microphysics of ice clouds is extremely complex. It is impossible to represent ice crystals in a GCM without any simplifications. However, the simplifications of microphysics may introduce biases into or reduce the variabilities of radiative properties of ice clouds and therefore result in an unrealistic climate. Some recent studies show that the short wave optical properties of ice clouds are highly sensitive to the habit distribution, inclusions and surface roughness of ice crystals. A failure in accounting for these microphysical properties may cause substantial errors in the remote sensing of optical properties of ice clouds. Unfortunately, none of these microphysical properties of ice crystals are accounted for in the community atmospheric model (CAM3.0). It has motivated us to investigate the sensitivity of radiative properties of ice clouds in CAM to these potentially important microphysical properties. Three new parameterization schemes have been developed and evaluated in the single column version of CAM. The impact of the new schemes on model climate will be studied and presented.