# Access Operators: **Index Seeks** > [!Note] > An Index Seek traverses a B-Tree to jump straight to matching keys, making it far more efficient than a Scan on both clustered and non-clustered indexes. **B-Tree structure.** A B-Tree is the ordered storage structure behind every SQL Server index. Its branching design supports logarithmic-time **searches**, **sequential reads**, and **updates** by guiding the engine from the root through intermediate pages down to the leaf level that holds either full rows (clustered) or key/locator pairs (non-clustered). **Clustered index seeks.** Because the [[Clustered Index]] stores the entire table in key order, an Index Seek on it can position directly on the first qualifying key and stream contiguous leaf pages to satisfy range predicates with minimal I/O. **Non-clustered index seeks.** A [[Non-Clustered Index]] holds only the indexed columns plus a row locator. An Index Seek here pinpoints matching keys quickly; if extra columns are needed, the engine performs a lookup to the clustered index or heap. Seeks on well-chosen non-clustered indexes are therefore ideal for highly selective filters. **Efficiency compared with scans.** Where a Scan touches every leaf page, a Seek touches just the few levels of the B-Tree plus the qualifying leaf pages—reducing disk reads, CPU work, and memory pressure. The optimizer therefore prefers Seeks whenever an appropriate index and selective predicate exist. --- ## References - Korotkevitch, D. (2022). _SQL Server advanced troubleshooting and performance tuning: Best practices and techniques_. O’Reilly Media. - Nevarez, B. (2022). _SQL Server query tuning and optimization: Optimize Microsoft SQL Server 2022 queries and applications_. Packt Publishing.