A critical analysis of the methods and results of estimating the optical thickness of the dust component in the Martian atmosphere tau (o), the particle size r(o), and the imaginary part of the refractive index n(i) has shown the following. (1) Observational data on the brightness distribution over the Martian disk as well as the phase dependences of diffusely reflected light and the azimuthal dependences of diffusely transmitted light are most appropriate to use only for verifying the reliability of the aerosol parameters determined by other methods. (2) If the morning and evening fogs in the atmosphere are disregarded, the Bouguer-Lambert-Beer method used to analyze the solar-brightness attenuation measured on the planetary surface yields overestimated extraatmospheric solar intensity I-o and atmospheric optical depth tau (o). At the Viking 1 landing site, I-o and tau (o) could be overestimated by a factor of 1.7 and by 0.35, respectively. (3) The aerosol size determined by analyzing measurements of the azimuthal dependences for the Martian sky brightness at low elevations of the Sun most likely corresponds to the fog particles. (4) If overestimated values of I-o were used to standardize the observations of the solar radiation transmitted by the Martian atmosphere, then n(i) were also overestimated; using overestimated tau (o) also affected the reliability of the latter. (5) The problem of reliability of the available tau (o) and r(o) estimates for periods of high atmospheric transparency is yet to be solved. For the highest activity of the dust storm in 1971, it was found that 4.5 less than or equal to r(o) less than or equal to 7.5 mum for the lognormal particle size distribution with sigma (2) = 0.2 and the optical thickness of a dust cloud tau (o) greater than or equal to 15. (6) The spectral values of the apparent albedo of Mars measured in October 1971 at a phase angle of 42 degrees in the spectral range 0.250 less than or equal to lambda less than or equal to 0.717 allowed the imaginary part of the refractive index to be estimated in terms of a model of a dust cloud composed of spherical particles with the lognormal size distribution with r(o) = 4.5 mum and sigma (2) = 0.2.