Java With Us

<code>LinkedHashMap</code> in Java — Insertion-Ordered Map

Updated on 2026-04-21 Reviewed by the Java With Us editorial team

LinkedHashMap is a HashMap that additionally maintains a doubly-linked list over its entries. Iteration visits entries in insertion order by default, or access order if you enable it — the building block of a simple LRU cache.

Basics

var cfg = new LinkedHashMap<String, String>();
cfg.put("host", "db.local");
cfg.put("port", "5432");
cfg.put("user", "admin");

cfg.forEach((k, v) -> System.out.println(k + "=" + v));
// host=db.local
// port=5432
// user=admin   ← always in insertion order

Access order — LRU cache

// initialCapacity, loadFactor, accessOrder
var cache = new LinkedHashMap<String, byte[]>(16, 0.75f, true) {
    @Override
    protected boolean removeEldestEntry(Map.Entry<String, byte[]> eldest) {
        return size() > 100;          // bounded to 100 entries
    }
};

cache.put("a", dataA);
cache.put("b", dataB);
cache.get("a");                        // moves 'a' to most recently used
cache.put("c", dataC);
// When size exceeds 100, removeEldestEntry drops the LRU automatically

Performance

Same O(1) average as HashMap. Extra memory: two pointers per entry (prev/next) and a slightly larger iteration overhead — typically <10%.

Common mistakes

  • Using HashMap then being surprised by order — iteration isn't guaranteed. Switch to LinkedHashMap.
  • Expecting a thread-safe LRU — LinkedHashMap isn't. Wrap in Collections.synchronizedMap, or use Caffeine/Guava for production caches.
  • Mixing access order and iteration — a get during iteration modifies the order, which can confuse the iterator.

Related

Pillar: Java collections. Siblings: HashMap, TreeMap.

Related articles