In order to profoundly understand abilities of two classifiers—step wise cluster and fuzzy cluster in the cloud classification techniques, both EOS/ MODIS and GMS5 data set are used, spectral or textural features are drawn from samples randomly to identify various cloud/surface. The results show that the st epwise cluster gives higher accuracies than fuzzy classifier on the whole. With rega rds to discriminating diverse cloud/surfaces, fuzzy cluster demonstrates its hig her accuracies than stepwise cluster on the classes having similar characteristi cs such as stratus, cumulostratus and cumulus; while stepwise cluster has better capabilities of distinguishing cumulonimbus and surfaces. As far as misclassifi cation of cloud/surfaces, fuzzy cluster tends to show lower accuracies in more m isclassified classes.