Description: By covering nearly 70% of our planet’s area, clouds exert enormous influence on the radiation, precipitation and energy budget. Coupling of clouds with other atmospheric components and associated processes occur over a wide range of spatio-temporal scales, thus making it difficult to observe and model them. The largest uncertainties in climate models are usually associated with clouds. In the last few decades, the satellite-based observations have gradually improved our understanding of clouds, while providing constraints on the climate models. Retrieving the historical, satellite-based observations of clouds from the heritage sensors has always been very challenging. A stratospheric progress has been made in the last 10-15 years in cloud remote sensing, mainly on two fronts. The introduction of active lidar and radar sensors in space has undeniably been a game-changer. The data from the CALIPSO and CloudSat satellites have enabled near-global training, tremendous improvements and rigorous validations of algorithms used to derive cloud properties from the heritage sensors. On the second front, there have been remarkable improvements in the (inter)calibration of raw data and the treatment of spurious quality artifacts, resulting in the high-quality fundamental climate data records. This contribution presents a comprehensive assessment of the latest versions of these four long-term cloud climate data records, namely CLARA-A3, ESA Cloud CCI, ISCCP and PATMOS-X. We will investigate the robust trends in global cloudiness, cloud subtypes and the commonalities across their spatio-temporal features. We debate whether or not providing the observational ensemble mean of cloud properties is the way forward.