The paper determines whether the earliest published indication of changes in GNP provide information which is useful in the interpretation of economic trends or whether subsequent revisions reveal that the true movements were radically different from those initially reported. An analysis of the provisional and advance data indicated that the initial changes differed from the final changes but still approximated the true pattern of the movements. Some of the data revisions, however, where unsuccessful. A further analysis of the temporal discrepancies in the yearly changes of one series indicates that there is not a wide degree of uncertainty about the actual movements. The combined evidence indicates that the early data are useful for economic analysis.