How to Create an Empty List in Python

You can create an empty list in Python with a pair of square brackets, items = [], or with items = list(). Both give you a new list with no elements. I use [], because it’s shorter and faster:

cart = []         # empty list
cart = list()     # same result

Python 3.12.5 produced the outputs below, and mypy checked the type-hint example.

You’ll see how the two forms compare, which one is faster, how to type an empty list, and the two traps that make “empty” lists share data when you don’t expect it.

Create an empty list with [] or list()

Each [] or list() makes a separate list. They’re equal while both are empty, but they’re never the same object:

cart = []
wishlist = list()

print(cart, wishlist, type(cart).__name__)
print("length:", len(cart), "| truthy?", bool(cart))
print("equal:", cart == wishlist, "| same object:", cart is wishlist)

cart.append("Coffee beans")
print("after append:", cart, "| wishlist still:", wishlist)

Output:

[] [] list
length: 0 | truthy? False
equal: True | same object: False
after append: ['Coffee beans'] | wishlist still: []

Once you have the list, filling it is usually a loop with append(); I cover that in adding elements to an empty list in Python.

Is [] faster than list() in Python?

Yes. [] is built straight into the bytecode, while list() has to look up the name and call it. The gap is tiny per call, but it’s measurable:

import timeit

n = 5_000_000
brackets = timeit.timeit("[]", number=n)
function = timeit.timeit("list()", number=n)
print(f"[]     {brackets:.3f} s for {n:,} lists")
print(f"list() {function:.3f} s for {n:,} lists")
print(f"[] is about {function / brackets:.1f}x faster")

print(list("TX"), list(range(3)), list({"a": 1}))   # list() is for converting things

Output:

[]     0.078 s for 5,000,000 lists
list() 0.141 s for 5,000,000 lists
[] is about 1.8x faster
['T', 'X'] [0, 1, 2] ['a']

So use [] for a new empty list and list(x) for converting a string, range or other iterable. More on making lists with starting values is in creating a list in Python.

Why do my empty lists share the same items?

Two common patterns look like they make fresh empty lists but don’t. A default argument like cart=[] is created once, and [[]] * 3 repeats one list three times:

def add_item(item, cart=[]):              # the default list is created ONCE
    cart.append(item)
    return cart

print(add_item("Mug"))
print(add_item("Filters"))               # still holds "Mug" from the first call!

def add_item_fixed(item, cart=None):
    if cart is None:
        cart = []                         # a fresh empty list every call
    cart.append(item)
    return cart

print(add_item_fixed("Mug"), add_item_fixed("Filters"))

rows = [[]] * 3                           # three references to ONE list
rows[0].append("Austin")
print("[[]] * 3:", rows)

rows = [[] for _ in range(3)]             # three separate empty lists
rows[0].append("Austin")
print("comprehension:", rows)

Output:

['Mug']
['Mug', 'Filters']
['Mug'] ['Filters']
[[]] * 3: [['Austin'], ['Austin'], ['Austin']]
comprehension: [['Austin'], [], []]

Use None as the default and create the list inside the function. For lists of empty lists, use a comprehension; there are more patterns in initializing a list of size N in Python.

Type hints for an empty list

An empty list doesn’t tell a type checker what it will hold, so annotate it. Then tools like mypy catch the wrong type going in, even though Python itself runs the code anyway:

orders: list[str] = []          # an empty list that should only hold strings
totals: list[float] = []

orders.append("A-1009")
totals.append(49.99)
orders.append(1010)             # mypy flags this before you run anything
print(orders, totals)

python -m mypy typed_lists.py
python typed_lists.py

Output:

typed_lists.py:6: error: Argument 1 to "append" of "list" has incompatible type "int"; expected "str"  [arg-type]
Found 1 error in 1 file (checked 1 source file)
['A-1009', 1010] [49.99]

Use list[str] on Python 3.9 and later; older code imports List from typing. If you’re unsure what a variable holds at runtime, checking whether a list is empty and isinstance() help.

Make an existing list empty again

Assigning [] creates a new list for that one name. clear() empties the list itself, so every other reference sees it too:

cart = ["Coffee beans", "Mug"]
saved = cart                          # another name for the same list

cart = []                             # makes a NEW empty list for 'cart' only
print("cart:", cart, "| saved:", saved)

cart = saved
cart.clear()                          # empties the list everyone shares
print("cart:", cart, "| saved:", saved)

if not cart:
    print("The cart is empty.")

Output:

cart: [] | saved: ['Coffee beans', 'Mug']
cart: [] | saved: []
The cart is empty.

Pick based on whether other code should keep the old items. I compare more options in clearing a list in Python.

Empty list options at a glance

Goal	Use
New empty list	items = []
Convert an iterable to a list	list(iterable)
Empty list as a default argument	items=None, then items = [] inside
Several separate empty lists	[[] for _ in range(n)]
Empty a list everyone shares	items.clear()

Here are more list guides to read next:

• Add an element to the beginning of a list
• Remove duplicates from a list in Python
• Flatten a list of lists in Python
• Copy elements from one list to another

Frequently asked questions