We analyse one year of cloud data from the ISCCP C and D datasets. The two datasets differ by their retrieval algorithms and their definitions of the cloud types defined from the cloud top pressure and cloud optical depth. The differences between the two datasets are first described in terms of the total cloud cover, as well as its repartition in low, middle, and high level cloudiness. We also project the ISCCP C cloud classes into the ISCCP D cloud types to circumvent the problem of different cloud type definitions in the two datasets. The differences between the two datasets are then also investigated in terms of the most frequent cloud type. ¿ 1998 American Geophysical Union