How do I rename columns in a pandas DataFrame?

The short version

Use rename(columns={...}) with a dictionary mapping old names to new ones when you want to change a few columns. To replace every name at once, assign a list to df.columns. Like most pandas methods, rename returns a new frame, so reassign the result.

Renaming selected columns

import pandas as pd

df = pd.DataFrame({"Prod Name": ["pen"], "Unit Price": [2]})

df = df.rename(columns={"Prod Name": "product", "Unit Price": "price"})
print(df.columns.tolist())   # ['product', 'price']

Only the columns you name are changed, and any name not in the dictionary is left alone, which makes this safe on a wide frame.

Replacing all names at once

If you want to set every column name, assign a list of the right length.

df.columns = ["product", "price"]

This is quick, but the list must match the column count and order exactly, so it is easy to misalign on a wide frame. Prefer the dictionary form when you only need to fix a few.

Cleaning messy headers in bulk

Real files often arrive with mixed case and spaces. A small transformation tidies them all consistently.

df.columns = (
    df.columns
    .str.strip()
    .str.lower()
    .str.replace(" ", "_")
)

This lowercases, trims stray spaces and swaps spaces for underscores, giving predictable names like product and unit_price that are easy to type and reference.

The reassign reminder

This line looks like it renames but changes nothing, because the result is discarded:

df.rename(columns={"price": "cost"})   # result thrown away

Either reassign to df, or pass inplace=True. Reassigning is usually clearer and works naturally in a chain of steps.

Renaming the index

The same method renames index labels through the index argument, which is useful after grouping when the index holds meaningful values.

df.rename(index={0: "first_row"})