A technique is considered for retrieving the spatial distributions of respirable fractions of aerosol in the lower atmosphere on the basis of multifrequency lidar sounding data without the use of additional aerosol optical and microphysical parameters along a sounding path. For this purpose, it is suggested to replace the spectral values of the aerosol extinction coefficient involved in lidar equations with the linearly independent parameters of their approximation, and retrieve the spatial distributions of these parameters from the numerical solution of the set of equations composed of all wavelength-time lidar signal samples. As a result, the number of unknowns in the set of equations to be solved is significantly reduced, and its matrix becomes overdetermined, which can be used for selection of physically reasonable values of the aerosol backscattering phase function at the lidar operating wavelengths. An assumption that there are two segments at the sounding path with similar aerosol extinction coefficient profiles is used to determine the lidar calibration constants. An algorithm is suggested for the search for these segments by the wavelength-time structure of a lidar signal. The inverse problem of aerosol light scattering is solved on the basis of stable regression relations between the concentrations of respirable aerosol fractions and approximation parameters of the aerosol extinction spectrum. The stability of the technique developed to the calibration errors and spatial variations in the aerosol backscattering phase function is shown in numerical experiments on laser sounding of aerosol.