The manuscript “Retrieval of stratospheric aerosol extinction coefficients from OMPS-LP measurements” by Rozanov et al. presents a new retrieval algorithm to obtain vertical profiles of the aerosol extinction coefficient. The main claim of the paper is that by avoiding altitude normalization, the algorithm becomes almost completely independent of the “prior aerosol extinction profile.” However, in my view, the authors did not provide sufficient evidence to prove this point. Particularly, in Section 5, the authors wrote that uncertainties about the aerosol concentration at the normalization altitude would lead to a strong sensitivity to the a priori extinction profile across the entire vertical range. Figure 3 shows results for two algorithms (V1.0.9 and V2.1) and concludes that by removing altitude normalization in V2.1, the retrieved profiles become almost insensitive to a priori. However, I am afraid that the authors are comparing apples and oranges here. We (readers) do not know if the two algorithms use the same L1 data or different data because the differences in the magnitude of retrieved aerosol extinction coefficients are quite large between the two algorithms, as shown in Figure 5. The authors do not describe all the algorithmic differences between the two algorithms to convince the reader that the changes they see in Figure 3 are caused by the normalization at higher altitude.