Home › Python

TypeError: 'dict_keys' object is not subscriptable (Python) — how to fix

Traceback (most recent call last):
  File "app.py", line 4, in <module>
    first = d.keys()[0]
TypeError: 'dict_keys' object is not subscriptable
TypeError: 'dict_values' object is not subscriptable
TypeError: 'dict_items' object is not subscriptable

Quick fix: Wrap the view in list() before indexing: list(d.keys())[0], or get the first key without building a list: next(iter(d)). For values use next(iter(d.values())). If you only need to check membership, use key in d instead of indexing.

What it means

In Python 2, d.keys() returned a list. In Python 3 it returns a view (dict_keys), a live, set-like window onto the dictionary. Views support iteration, len(), in, and set operations (&, |, -), but not indexing or slicing. So code like this fails:

d = {"a": 1, "b": 2}
d.keys()[0]        # TypeError: 'dict_keys' object is not subscriptable
d.values()[-1]     # TypeError: 'dict_values' object is not subscriptable
d.items()[:2]      # TypeError: 'dict_items' object is not subscriptable

This is very common when running old Python 2 code or tutorials on Python 3.

Fix 1: convert the view to a list

keys = list(d.keys())
keys[0]           # 'a'
list(d.values())[-1]
list(d.items())[:2]

list(d) is the same as list(d.keys()). Since Python 3.7, dicts keep insertion order, so index 0 is the first key you inserted.

Fix 2: get just the first item without copying

Building a list of a huge dict just to read one element is wasteful. Use an iterator:

first_key = next(iter(d))
first_value = next(iter(d.values()))
first_key, first_value = next(iter(d.items()))

On an empty dict next() raises StopIteration; pass a default to avoid that:

first_key = next(iter(d), None)

For the first N items, use itertools.islice:

from itertools import islice
first_three = list(islice(d.items(), 3))

Fix 3: you didn't actually need an index

Often the index was only used to loop or test membership. Use the view directly:

for key, value in d.items():
    print(key, value)

if "a" in d:          # fast, no .keys() needed
    ...

common = d.keys() & other.keys()   # set operations work on views

Fix 4: random or sorted access

import random
random.choice(list(d))          # random key

sorted(d)[0]                    # alphabetically first key
max(d, key=d.get)               # key with the largest value

The same error appears with anything dict-like, e.g. json.load(...).keys(), os.environ.keys(), or collections.OrderedDict. The fix is the same: list(...) or next(iter(...)).

If you're porting Python 2 code, also watch for iteritems() (see AttributeError: 'dict' object has no attribute 'iteritems') and print statements.

Last updated 2026-11-12.