Identifying cloud regimes and their role in the climate system can serve a multitude of purposes, ranging from a better understanding of clouds to guiding field experiments to improving the representation of clouds in models. This study describes early results in identifying cloud regimes from ISCCP data using cluster analysis. A simple algorithm for cloud regime identification is introduced and applied to data in the Tropical Western Pacific region. Four major cloud regimes, namely a shallow cumulus regime, a transparent isolated cirrus regime, thick cirrus with convection and a deep and probably organized convective regime are identified and their frequency of occurrence is quantified. The use of the regime information for various applications is discussed and the use of regime classifications for representativeness studies is presented using the ARM TWP sites as an example.