Temp Head

Temp Head Technique

This is a very common linked list technique as it typically saves you creating special edge condition logic in order to operate on the head of a linked list with some algorithms. This technique only involves creating one extra pointer, the temporary head, that will point to your final answer or list that you will return. This technique is much easier to demonstrate with an example.

Note: Historically, “dumb” was used as a reference to a member of the deaf community. It has negative connotations as a derogatory term. Why do we use the term “temp head” instead of “dummy head”?

Words Matter. We want to call out these terms to increase awareness and ensure that we are inclusive in our communities. If you notice other words or behaviors that are not inclusive, please let a CodePath team member know or email

Check out these resources to learn more:

Delete Node

Say you are asked to delete a node in a linked list given the value of the node you want to delete. Furthermore you are told that you can assume the values are unique.

For example:

Input: 1->2->4, 2
Output: 1->4
class Node(object):
    def __init__(self, v):
        self.val = v = None
def delete_node(head, val):
    d = Node("temp") # 1 = head
    p = d
    c = head
    while c:
        if c.val == val:
   = # 2
            return   # 3
        p = c
        c =
    return # 4 
  1. The creation of the temporary head, in this case it is initialized to point to the original list.
  2. Because we created the temporary head we don't have to treat deleting the head of the original list any different from other elements in the list.
  3. The temporary head points to the answer list so we simply return the next node as the head of the answer.
  4. If we don't happen to find the item setting the temporary head to the original list makes it still point to the correct answer.


Treating the temporary head with the invariant that it is always pointing to the current correct answer makes dealing with edge cases in linked lists a lot easier and can be found in a number of elegant solutions.


Fork me on GitHub