What are the differences between a HashMap and a Hashtable in Java?

Note, that a lot of the answers state that Hashtable is synchronised. In practice this buys you very little. The synchronization is on the accessor / mutator methods will stop two threads adding or removing from the map concurrently, but in the real world you will often need additional synchronisation.

A very common idiom is to "check then put" — i.e. look for an entry in the Map, and add it if it does not already exist. This is not in any way an atomic operation whether you use Hashtable or HashMap.

An equivalently synchronised HashMap can be obtained by:

Collections.synchronizedMap(myMap);

But to correctly implement this logic you need additional synchronisation of the form:

synchronized(myMap) {
    if (!myMap.containsKey("tomato"))
        myMap.put("tomato", "red");
}

Even iterating over a Hashtable's entries (or a HashMap obtained by Collections.synchronizedMap) is not thread safe unless you also guard the Map from being modified through additional synchronization.

Implementations of the ConcurrentMap interface (for example ConcurrentHashMap) solve some of this by including thread safe check-then-act semantics such as:

ConcurrentMap.putIfAbsent(key, value);

Hashtable is considered legacy code. There's nothing about Hashtable that can't be done using HashMap or derivations of HashMap, so for new code, I don't see any justification for going back to Hashtable.

This question is often asked in interview to check whether candidate understands correct usage of collection classes and is aware of alternative solutions available.

1. The HashMap class is roughly equivalent to Hashtable, except that it is non synchronized and permits nulls. (HashMap allows null values as key and value whereas Hashtable doesn't allow nulls).
2. HashMap does not guarantee that the order of the map will remain constant over time.
3. HashMap is non synchronized whereas Hashtable is synchronized.
4. Iterator in the HashMap is fail-safe while the enumerator for the Hashtable is not and throw ConcurrentModificationException if any other Thread modifies the map structurally by adding or removing any element except Iterator's own remove() method. But this is not a guaranteed behavior and will be done by JVM on best effort.

Note on Some Important Terms:

1. Synchronized means only one thread can modify a hash table at one point of time. Basically, it means that any thread before performing an update on a Hashtable will have to acquire a lock on the object while others will wait for lock to be released.
2. Fail-safe is relevant within the context of iterators. If an iterator has been created on a collection object and some other thread tries to modify the collection object "structurally", a concurrent modification exception will be thrown. It is possible for other threads though to invoke set method since it doesn't modify the collection "structurally". However, if prior to calling set, the collection has been modified structurally, IllegalArgumentException will be thrown.
3. Structurally modification means deleting or inserting element which could effectively change the structure of map.

HashMap can be synchronized by

Map m = Collections.synchronizeMap(hashMap);

Map provides Collection views instead of direct support for iteration via Enumeration objects. Collection views greatly enhance the expressiveness of the interface, as discussed later in this section. Map allows you to iterate over keys, values, or key-value pairs; Hashtable does not provide the third option. Map provides a safe way to remove entries in the midst of iteration; Hashtable did not. Finally, Map fixes a minor deficiency in the Hashtable interface. Hashtable has a method called contains, which returns true if the Hashtable contains a given value. Given its name, you'd expect this method to return true if the Hashtable contained a given key, because the key is the primary access mechanism for a Hashtable. The Map interface eliminates this source of confusion by renaming the method containsValue. Also, this improves the interface's consistency — containsValue parallels containsKey.

The Map Interface

HashMap: An implementation of the Map interface that uses hash codes to index an array. Hashtable: Hi, 1998 called. They want their collections API back.

Seriously though, you're better off staying away from Hashtable altogether. For single-threaded apps, you don't need the extra overhead of synchronisation. For highly concurrent apps, the paranoid synchronisation might lead to starvation, deadlocks, or unnecessary garbage collection pauses. Like Tim Howland pointed out, you might use ConcurrentHashMap instead.

Keep in mind that HashTable was legacy class before Java Collections Framework (JCF) was introduced and was later retrofitted to implement the Map interface. So was Vector and Stack.

Therefore, always stay away from them in new code since there always better alternative in the JCF as others had pointed out.

Here is the Java collection cheat sheet that you will find useful. Notice the gray block contains the legacy class HashTable,Vector and Stack.