Sort a dictionary by keys or values in Python

Sort mapping.items() to get an ordered list of key-value pairs. If you need a dictionary, pass that list to dict(): in Python 3.7 and later, dictionaries preserve insertion order.

Sort by key

values = {"c": 3, "a": 1, "b": 2}
items = sorted(values.items(), key=lambda item: item[0])
print(items)  # [('a', 1), ('b', 2), ('c', 3)]

ordered = dict(items)
print(ordered)  # {'a': 1, 'b': 2, 'c': 3}

For ordinary comparable keys, dict(sorted(values.items())) is a shorter form: tuples are compared starting with their first element. If you need only the keys, use sorted(values).

Sort by value

values = {"a": 1, "c": 3, "b": 2}
ascending = dict(sorted(values.items(), key=lambda item: item[1]))
descending = dict(sorted(
    values.items(), key=lambda item: item[1], reverse=True
))
print(list(descending))  # ['c', 'b', 'a']

The original dictionary stays unchanged. If you keep the sorted list of pairs instead, you can iterate in order without rebuilding a dictionary.

Choose a tie-breaker

Sorting is stable: entries with equal values retain their original insertion order. To order equal values by key instead, use a tuple key:

values = {"b": 2, "a": 2, "c": 1}
items = sorted(values.items(), key=lambda item: (item[1], item[0]))
print(items)  # [('c', 1), ('a', 2), ('b', 2)]

The key fields must be mutually comparable. A mixture of numeric and string values can raise TypeError; convert or validate them according to the intended semantics.

A dictionary does not stay sorted automatically

Insertion order is not a sorted-index data structure. Adding a new key puts it at the end, regardless of where it belongs in the earlier sort. Replacing an existing value does not move that key. Sort again when you need a fresh ordering, and use a suitable specialized data structure for repeated range queries.

FAQ

Do I still need OrderedDict? An ordinary dictionary is sufficient when you only need insertion order. OrderedDict remains useful for explicit reordering with move_to_end, popping from either end, and order-sensitive equality between two OrderedDict instances.

Does reverse=True change tie behavior? No. Equal sort keys retain their relative order. Use an explicit secondary key if you want a deterministic tie-breaker independent of insertion history.

References

• Python built-in types
• Python collection types
• Python sorting guide