The distortions of the inverted lidar signals can be caused by (i) the constant offset that remains in the backscatter signal after removing the background component, (ii) the multiplicative distortion component, which level is related with the lidar signal intensity, and (iii) the signal noise in the wide wavelength spectra; the latter includes lowfrequency components, which do not obey common random-noise statistics. These distortions, even minor, may yield significant distortions in the retrieved outputs obtained by the inversion of the lidar signal. Implicit and explicit premises and assumptions required for any solution of the lidar equation are additional sources of the uncertainty in the inversion results. There is no reliable way for checking whether used assumptions are valid, therefore, the lidar signal inversion can yield significantly biased results. As a result, instead some statistically mean profile of the atmospheric parameter of interest with the corresponding probability function, one obtains some qualitative estimate of the profile with unknown uncertainty, which depends on the validity of used assumptions. It means that lidar profiling is not a measurement but a result of some simulation based on past observations.