You can use the following basic syntax to add or subtract time to a datetime in pandas:

#add time to datetime
df['new_datetime'] = df['my_datetime'] + pd.Timedelta(hours=5, minutes=10, seconds=3)

#subtract time from datetime
df['new_datetime'] = df['my_datetime'] - pd.Timedelta(hours=5, minutes=10, seconds=3) 

The following example shows how to use this syntax in practice.

Example: Add/Subtract Time to Datetime in Pandas

Suppose we have the following pandas DataFrame that shows the sales made by some store during 10 different datetimes:

import pandas as pd

#create DataFrame
df = pd.DataFrame({'time': pd.date_range('2022-01-01', periods=10),
                   'sales': [14, 22, 25, 29, 31, 10, 12, 8, 22, 25]})

#view DataFrame
print(df)

        time  sales
0 2022-01-01     14
1 2022-01-02     22
2 2022-01-03     25
3 2022-01-04     29
4 2022-01-05     31
5 2022-01-06     10
6 2022-01-07     12
7 2022-01-08      8
8 2022-01-09     22
9 2022-01-10     25

We can use the pandas Timedelta function to add 5 hours, 10 minutes, and 3 seconds to each datetime value in the “time” column:

#create new column that contains time + 5 hours, 10 minutes, 3 seconds
df['time_plus_some'] = df['time'] + pd.Timedelta(hours=5, minutes=10, seconds=3)

#view updated DataFrame
print(df)

        time  sales      time_plus_some
0 2022-01-01     14 2022-01-01 05:10:03
1 2022-01-02     22 2022-01-02 05:10:03
2 2022-01-03     25 2022-01-03 05:10:03
3 2022-01-04     29 2022-01-04 05:10:03
4 2022-01-05     31 2022-01-05 05:10:03
5 2022-01-06     10 2022-01-06 05:10:03
6 2022-01-07     12 2022-01-07 05:10:03
7 2022-01-08      8 2022-01-08 05:10:03
8 2022-01-09     22 2022-01-09 05:10:03
9 2022-01-10     25 2022-01-10 05:10:03

And we can just as easily create a new column that subtracts 5 hours, 10 minutes, and 3 seconds from each datetime value in the “time” column:

#create new column that contains time - 5 hours, 10 minutes, 3 seconds
df['time_minus_some'] = df['time'] - pd.Timedelta(hours=5, minutes=10, seconds=3)

#view updated DataFrame
print(df)

        time  sales     time_minus_some
0 2022-01-01     14 2021-12-31 18:49:57
1 2022-01-02     22 2022-01-01 18:49:57
2 2022-01-03     25 2022-01-02 18:49:57
3 2022-01-04     29 2022-01-03 18:49:57
4 2022-01-05     31 2022-01-04 18:49:57
5 2022-01-06     10 2022-01-05 18:49:57
6 2022-01-07     12 2022-01-06 18:49:57
7 2022-01-08      8 2022-01-07 18:49:57
8 2022-01-09     22 2022-01-08 18:49:57
9 2022-01-10     25 2022-01-09 18:49:57

Note #1: In these examples we used a specific number of hours, minutes, and seconds, but you can also use only one of these units if you’d like. For example, you can specify pd.Timedelta(hours=5) to simply add five hours to a datetime value.

Note #2: You can find the complete documentation for the pandas Timedelta function .

The following tutorials explain how to perform other common tasks in pandas: