The Monte Carlo numerical experiment has been used to study the efficiency of rotational-vibrational Raman spectroscopy during laser sensing of temperature and humidity in the atmosphere. In the real atmosphere, the main source of active noise, limiting the potential capabilities of a lidar, is multiple scattering of a laser signal by aerosol and cloud particles. Under the lower- and middle-level overcast conditions, the Raman lidar is inapplicable. It is an urgent problem to estimate the limits of the Raman sensing applicability in the presence of aerosol inversions and invisible upper-level cirrus clouds. In this paper, we estimate a possible shift of vertical profiles of temperature, humidity, and the ratio of H2O vapor mixture due to the multiple scattering noise in the recording channel for cases of ground-based and orbital sensing. The boundary conditions of the problem correspond to those of the most efficient active lidars of the European lidar network, using the signals of purely rotational and rotational-vibrational Raman scattering induced by the pulsed Nd:YAG laser radiation at a wavelength of 532.25 nm. The estimates confirm the promises of applying the methods of rotational-vibrational Raman spectroscopy to sensing the temperature in a height range 2-20 km, while the errors in estimation of water vapor profiles achieve 10-15%.