Comparability among measurements from Automatic Lidar-Ceilometers (ALCs) in heterogeneous measurement networks (such as EUMETNET E-PROFILE) strongly depends on the accuracy of the adopted calibration procedures. Furthermore, the retrieval of higher-level products such as aerosol extinction and mass concentration is based on attenuated backscatter; hence, large uncertainties in the calibration constant strongly hamper the use of ALC products from an end-user perspective. Several calibration techniques have been developed in the last decades, notably the one based on molecular/Rayleigh scattering in an aerosol-free layer of the atmosphere. One severe uncertainty is associated with a seasonal cycle found in the calibration values. This systematic variation is still of unknown origin but needs to be carefully addressed during the application of the calibration procedures. In this VMG, a statistical approach to the data is adopted to investigate the origin of the cycle by analyzing a large number of calibration datasets across Europe provided by the E-PROFILE network, covering a strong diversity of instrument models and ages, and environmental conditions. For the first time, the amplitudes of the cycle over such a large scale (involving 70 stations) are homogeneously compared using a single method. As a first step, the E-PROFILE calibration dataset is compared for a selection of sites to the results of the calibration methods adopted in the frame of other European initiatives, such as the DWD (Germany) and the Alicenet (Italy) ALC networks. No relevant systematic differences are found in the comparison of the calibration methods, with discrepancies mainly arising from single outliers attributed to the presence of elevated aerosol layers. The variations of the magnitude of the cycle found during the analysis among the European sites do not reveal any clear geographical pattern, e.g., as a function of latitude or other environmental characteristics of the sites. In the second part of the study, the temporal variations of the ALC calibration factor are correlated with some instrumental parameters in several sites. Significant similarities are found between the variation of the calibration factor and the following parameters: the laser pulse numbers, the detector sensitivity and the background noise. Finally, an on-going assessment to find a way to express the lidar constants only with housekeeping data showed first encouraging results. Progress on this topic will help networks develop operational calibration procedures.