Two layers of Saharan dust observed over Leipzig, Germany, in February and March 2021 were used to provide the first-ever lidar measurements of the dust lidar ratio (extinction-to-backscatter ratio) and linear depolarization ratio at all three classical lidar wavelengths (355, 532 and 1064 nm). The pure-dust conditions during the first event exhibit lidar ratios of 47 ± 8, 50 ± 5 and 69 ± 14 sr and particle linear depolarization ratios of 0.242 ± 0.024, 0.299 ± 0.018 and 0.206 ± 0.010 at wavelengths of 355, 532 and 1064 nm, respectively. The second, slightly polluted-dust case shows a similar spectral behavior of the lidar and depolarization ratio with values of the lidar ratio of 49 ± 4, 46 ± 5 and 57 ± 9 sr and the depolarization ratio of 0.174 ± 0.041, 0.298 ± 0.016 and 0.242 ± 0.007 at 355, 532 and 1064 nm, respectively. The results were compared with Aerosol Robotic Network (AERONET) version 3 (v3) inversion solutions and the Generalized Retrieval of Aerosol and Surface Properties (GRASP) at six and seven wavelengths. Both retrieval schemes make use of a spheroid shape model for mineral dust. The spectral slope of the lidar ratio from 532 to 1064 nm could be well reproduced by the AERONET and GRASP retrieval schemes. Higher lidar ratios in the UV were retrieved by AERONET and GRASP. The enhancement was probably caused by the influence of fine-mode pollution particles in the boundary layer which are included in the columnar photometer measurements. Significant differences between the measured and retrieved wavelength dependence of the particle linear depolarization ratio were found. The potential sources for these uncertainties are discussed.