Simulations of Titan's atmospheric transmission and surface reflectivity have been developed in order to estimate how Titan's atmosphere and surface properties could affect performances of the Cassini radar experiment. In this paper we present a selection of models for Titan's haze, vertical rain distribution, and surface composition implemented in our simulations. We collected dielectric constant values for the Cassini radar wavelength (∼ 2.2 cm) for materials of interest for Titan: liquid methane, liquid mixture of methane-ethane, water ice, and light hydrocarbon ices. Due to the lack of permittivity values for Titan's haze particles in the microwave range, we performed dielectric constant (ε r) measurements around 2.2 cm on tholins synthesized in laboratory. We obtained a real part of ε r in the range of 2-2.5 and a loss tangent between 10 −3 and 5 × 10 −2. By combining aerosol distribution models (with hypothetical condensation at low altitudes) to surface models, we find the following results: (1) Aerosol-only atmospheres should cause no loss and are essentially transparent for Cassini radar, as expected by former analysis. (2) However, if clouds are present, some atmospheric models generate significant attenuation that can reach −50 dB, well below the sensitivity threshold of the receiver. In such cases, a 13.78 GHz radar would not be able to measure echoes coming from the surface. We thus warn about possible risks of misinterpretation if a "wet atmosphere" is not taken into account. (3) Rough surface scattering leads to a typical response of ∼ −17 dB. These results will have important implications on future Cassini radar data analysis.< Réduire