Python Exception Handling: try/except/raise

Part 11 of the Python for Everyone track. Last updated: September 2026.

Exceptions aren't crashes to fear — they're Python's normal way of signaling "something unexpected happened." Master the four-block pattern and you'll write code that fails gracefully instead of mysteriously.

try / except / else / finally

try:
    age = int(input("Age: "))        # may raise ValueError
    print(f"Next year: {age + 1}")
except ValueError:
    print("Please enter a number!")
except (TypeError, OverflowError) as e:   # catch multiple, keep the object
    print(f"Unexpected problem: {e}")
else:
    print("No errors — runs only if try succeeded")    # rare but clean
finally:
    print("Always runs — cleanup goes here")

Order matters: put specific exceptions before general ones, and never use a bare except: — it swallows KeyboardInterrupt and hides bugs.

raise: throw your own

def withdraw(balance, amount):
    if amount < 0:
        raise ValueError("amount must be positive")
    if amount > balance:
        raise ValueError(f"insufficient funds: {balance}")
    return balance - amount

try:
    withdraw(100, 200)
except ValueError as e:
    print(e)   # insufficient funds: 100

raise ... from ... chains exceptions when translating low-level errors into domain errors.

Custom exceptions

class PaymentError(Exception):
    """Base for all payment failures in our app."""
    pass

class InsufficientFundsError(PaymentError):
    pass

try:
    raise InsufficientFundsError("balance $10, need $25")
except PaymentError as e:      # catches the subclass too
    print(f"Payment failed: {e}")

Subclassing Exception (not BaseException) keeps your errors catchable without trapping system exits.

EAFP: it's easier to ask forgiveness

# LBYL (look before you leap) — Java-style
if "key" in config:
    value = config["key"]

# EAFP — Pythonic
try:
    value = config["key"]
except KeyError:
    value = "default"

Python culture prefers EAFP (try first, handle the exception) over LBYL (check everything first) — it's faster in the common case and avoids race conditions.

Key takeaways

• try/except/else/finally — else runs on success, finally always runs.
• Catch specific exceptions; never bare except:.
• raise your own errors; subclass Exception for custom ones.
• Prefer EAFP over LBYL — it's the Pythonic way.

Next in this series: Python Decorators and Generators Explained.