An assessment on the performance of a dual-wavelength (13.6/35.0 GHz) precipitation radar to observe snow and rain from space is drawn based on a simplified model, so-called noncoalescence and nonbreakup model, for falling hydrometeors. The simulation, first, generates vertical profiles of dual-wavelength radar integral parameters using raindrop size distribution (DSD) data measured at ground. The simulation results are then compared with some typical observations. Despite many uncertainties and assumptions, the simulated reflectivity profiles have shown some degrees of consistency with observations. The results based on statistical analysis of the simulated effective reflectivity factors (Zes) at 13.6 (Ze14) and 35 (Ze35) GHz radar frequencies indicate that: (1) rain and dry snow signatures are separable, if Ze14 is above ~10 dBZ, (2) separation of rain and wet snow signatures is feasible, only if Ze14 is between ~25 and ~45 dBZ, while above ~45 dBZ (Ze14) rain and wet snow signatures are poorly separable, and (3) separation of Ze signatures for dry and wet snow is rarely feasible. Moreover, from the analysis of the maximum allowable fluctuation in the standard deviations in Zes for rain and snow (dry and wet), this study proposes rain/snow discrimination thresholds. The rain/snow separation aspect is also studied with different distributions of snow density. Also, this study proposes conventional type Ze14-R relationships for snowfall rate retrieval for dry and wet snow regions.