Range-Based Set Reconciliation (RBSR) is an algorithm for efficiently finding differences between large sets held by different peers. Instead of exchanging all item identifiers, each peer recursively compares hash summaries of ordered key ranges. When the hash of a range differs, the peers subdivide that range and compare smaller segments until they isolate the exact items that differ. This approach allows them to reconcile large datasets with minimal bandwidth use, as the cost grows with the number of differences, not the total set size.