How to Drop Rows by Condition in Pandas DataFrame

To drop rows by condition in pandas, use boolean indexing with df = df[condition] to keep rows that meet the condition or use df.drop() with row indices that meet the condition. This lets you filter out unwanted rows easily.

Syntax

Use boolean indexing to keep rows that meet a condition:

df = df[condition]

Or use drop() with indices of rows to remove:

df = df.drop(indices)

Here, condition is a boolean expression on DataFrame columns, and indices are row labels to drop.

python

df = df[condition]
df = df.drop(indices)

Example

This example shows how to drop rows where the column 'Age' is less than 30.

python

import pandas as pd

data = {'Name': ['Alice', 'Bob', 'Charlie', 'David'],
        'Age': [25, 35, 30, 22]}
df = pd.DataFrame(data)

# Drop rows where Age is less than 30
condition = df['Age'] >= 30
df_filtered = df[condition]

print(df_filtered)

Output

      Name  Age
1      Bob   35
2  Charlie   30

Common Pitfalls

One common mistake is trying to use drop() with a boolean condition directly, which does not work. Instead, you must get the indices of rows to drop.

Also, forgetting to assign the result back to the DataFrame means the original data is unchanged.

python

import pandas as pd

data = {'Name': ['Alice', 'Bob', 'Charlie', 'David'],
        'Age': [25, 35, 30, 22]}
df = pd.DataFrame(data)

# Wrong: drop() does not accept boolean condition directly
# df.drop(df['Age'] < 30)  # This will raise an error

# Right: get indices where condition is True
indices_to_drop = df[df['Age'] < 30].index
df = df.drop(indices_to_drop)

print(df)

Output

      Name  Age
1      Bob   35
2  Charlie   30

Quick Reference

Method	Usage	Description
Boolean Indexing	df = df[condition]	Keep rows where condition is True
Drop by Index	df = df.drop(indices)	Remove rows by their index labels
Get Indices by Condition	indices = df[condition].index	Find row indices matching condition

Key Takeaways

Use boolean indexing to keep rows that meet your condition.

Use df.drop() with row indices to remove rows by condition.

Always assign the result back to your DataFrame to save changes.

You cannot pass a boolean condition directly to df.drop().

Get indices of rows to drop using df[condition].index.