In the present research, Iran was studied and zoned based on the extent and magnitude of foreshocks before the occurrence of earthquakes larger than magnitude 5 using a data-mining process. The aim of this research is to stress the importance of foreshock precursors for different zones of the country; therefore, these zones will be important for earthquake prediction research based on study of precursors. To conduct this research, foreshock precursors were introduced and then separated from the seismic database for the country considering reliable references during declustering operations. After preparing a foreshock database for the country, clustering operations were performed on it using the self-organizing map (SOM) and k-means methods. Using the silhouette index, it was found that the best classification of foreshocks was to classify them into six main clusters, and then group these clusters using Duncan and Tukey statistical tests for investigation in terms of magnitude. Finally, the terminal sequence of the Zagros–Makran Transition Zone was determined to be the main zone of the country in terms of number, relation, and magnitude of foreshocks before the occurrence of earthquakes of magnitude larger than 5. The Hormozgan region is completely located in this zone; i.e., foreshocks have a very close relation with large earthquakes, and most earthquakes in this region were accompanied by foreshocks of relatively high magnitude.