Python Lists vs Tuples vs Sets vs Dictionaries: When to Use Each

The Python list vs tuple vs set question comes up in every beginner project, and the answer is shorter than most tutorials make it. Each structure is built for one kind of question, and dictionaries cover a fourth.

Choose well in any Python list vs tuple vs set decision and your code gets faster and clearer. Choose badly and you end up de-duplicating lists by hand or scanning a million items to check one name.

This guide compares lists, tuples, sets and dictionaries side by side, shows real timings from a quick benchmark, and finishes with a decision table plus the mistakes testers and developers make most often. If you are new to the language, start with our Python tutorial for beginners.

Python list vs tuple vs set: quick comparison

Dictionaries keep insertion order since Python 3.7. Sets do not promise any order, so never rely on the order you see when printing one. The Python docs describe all four in the data structures tutorial.

Lists: the default for ordered, changeable data

Use a list when you need a sequence you will add to, remove from or sort. Test data rows, API results and queued tasks are typical.

users = ["asha", "ben", "asha"]
users.append("chen")
users.sort()
print(users)        # ['asha', 'asha', 'ben', 'chen']
print(users[0])     # asha

Lists are flexible, but that flexibility has a cost. Looking up a value with in checks items one by one, so it slows down as the list grows.

Tuples: a fixed record

A tuple says “these values belong together and will not change”. Coordinates, database rows and function return values fit well.

point = (12.5, 77.6)
lat, lon = point              # unpacking
def min_max(nums):
    return min(nums), max(nums)   # returns a tuple
low, high = min_max([4, 1, 9])

Because tuples cannot change, Python can use them as dictionary keys, provided every item inside is hashable too. A tuple holding a list is not.

Sets: uniqueness and fast membership

Use a set to remove duplicates or to ask “is this item present?” many times. Sets also support maths-style operations.

a = {"login", "search", "cart"}
b = {"cart", "checkout"}
print(a | b)    # union
print(a & b)    # intersection: {'cart'}
print(a - b)    # difference
print(len(set([3, 3, 1, 1, 2])))   # 3

Note that {} creates an empty dictionary. Use set() for an empty set. If you need an immutable set, use frozenset.

Dictionaries: lookup by key

If you often ask “what is the value for this name?”, use a dict.

prices = {"apple": 1.2, "pear": 0.9}
prices["kiwi"] = 1.5
print(prices.get("mango", "n/a"))   # n/a
for name, price in prices.items():
    print(name, price)

Dictionaries are the backbone of JSON handling, so they appear constantly in API work; see what is an API for the request side.

Python list vs tuple vs set speed: a real timing

I ran a worst-case lookup (the last item) on a collection of 100,000 integers using Python 3.13.6 on a laptop and timeit:

Your numbers will differ, but the gap is the point: hash-based lookup is roughly four orders of magnitude faster at this size. The official time complexity page lists the average cost as O(1) for sets and dicts and O(n) for lists and tuples. Memory differs too: on the same machine, a three-item tuple used 64 bytes versus 88 for a list, and 216 bytes for a set, so small tuples are cheaper to hold in large numbers.

Python list vs tuple vs set: which one should you use?

Common mistakes

• Using a list for membership tests inside a loop. Convert to a set first.
• Expecting set order. Sort the result if order matters.
• Mutable default arguments. def f(x=[]) shares one list across calls; use None instead.
• Putting a list inside a set. It raises TypeError: unhashable type; use a tuple.
• Editing a list while looping over it. Loop over a copy or build a new list.

When you test code built on a Python list vs tuple vs set choice, a parameterised test is a handy way to check them; our guide to writing test cases shows the format.

Final thoughts

For the Python list vs tuple vs set decision, ask what you do most: edit a sequence, keep a fixed record, check membership or look up by key. Start with a list, switch to a set or dict when lookups slow you down, and use tuples for data that should not change. That rule covers most real code.

Frequently asked questions

Keep learning