The validation of cloudiness in general circulation models has traditionally been done by comparing time-averaged quantities, such as monthly mean cloud cover, with available observations. This approach ignores the dynamical aspects of clouds such as their lifecycle characteristics, areal coverage, temporal duration, and internal variability. In this study, a new Lagrangian approach to the validation of modeled tropical cloudiness is explored. An automated cloud detection and tracking algorithm is used to observe and track overcast decks of cloud with scales from 300 km to greater than 3000 km in a consecutive set of half-hourly METEOSAT-5 images and the NCAR Community Climate Model (CCM3) during the winter monsoon season. Statistics are compiled and compared based on large ensembles of both observed clouds and modeled clouds whose properties, such as areal coverage, temporal duration and cloud-averaged precipitation rate, have been observed throughout the lifecycle of each cloud. The largest clouds in the satellite images are semi-permanent overcast decks of anvil and cirrus cloud, which are maintained by several regions of deep convection embedded within. Important internal variability occurs at timescales shorter than the lifetime of the overcast deck, including diurnal variability in both convective and cirrus cloud area. The presence of clouds in CCM3 is currently predicted according to relationships with convection and relative humidity. The evolution of these predictor quantities is tracked through the lifecycles of the modeled clouds and the ability of CCM3 to simulate clouds with the observed spatio-temporal variability is determined.