Based on the research, here's how Prefect handles your scenario: When you deploy flows with the same name from different files, Prefect will treat them as the same flow definition but with different deployments. This means: 1. When you execute the "test" deployment, it will run the flow implementation from

test_flow.py

2. When you execute the "prod" deployment, it will run the flow implementation from

prod_flow.py

Each deployment maintains its own reference to its source code location, so even though the flows share the same name, they execute their respective implementations correctly. However, this setup isn't recommended because: 1. It can lead to confusion in the UI since both deployments appear under the same flow 2. It makes it harder to track which implementation is actually being run 3. It could cause issues with flow run history and monitoring Better Practices: 1. Give your flows different names to clearly distinguish them:

python
# test_flow.py
@flow(name="test_my_flow")
def my_flow():
    ...

# prod_flow.py
@flow(name="prod_my_flow")
def my_flow():
    ...

2. Or keep the flows in the same file with clear naming:

python
@flow(name="my_flow_test")
def test_version():
    ...

@flow(name="my_flow_prod")
def prod_version():
    ...

This way, you'll have clear separation in the UI and better traceability of which version is running. For more information about deployment best practices, you can refer to the Prefect deployment documentation.