The sliding window pattern: how to spot it

Updated Oct 15, 2026

A sliding window keeps a contiguous range and its state (a sum, counts, distinct values) and updates them as the range moves, instead of recomputing every subarray. Fixed windows slide by one; variable windows grow on the right and shrink on the left.

When to reach for it

• The question is about a contiguous subarray or substring.
• It asks for the longest, shortest, or best range that meets a condition.
• The condition can be checked from state you can add to and remove from (a sum, character counts).
• A brute force restarts from every starting index.

left = 0
state = empty
best = initial
for right, x in enumerate(items):
    add(state, x)
    while invalid(state):
        remove(state, items[left])
        left += 1
    best = better(best, right - left + 1)
return best

4 LeetCode sliding window problems, easiest first

1. Maximum Average Subarray I easy

   Seed k cells, then slide. Max average is max sum over a fixed length.

2. Longest Substring Without Repeating Characters medium

   A repeat makes the window invalid. Shrink left until the character is unique again.

3. Minimum Size Subarray Sum medium

   Expand until the sum is enough, then shrink for the shortest cover.

4. Max Consecutive Ones III medium

   Zeros are the budget. Too many zeros: slide left until you are back under k.

Skills these problems use

Each one has a free lesson with worked examples and a drill.

• Fix the window length in advance
• Slide by dropping left and adding right
• Maintain the sum of the window
• Unique-character window shrinks on a repeat
• Maintain distinct count in the window
• Track the best valid length
• Shrink to the smallest valid window
• Track the best valid bounds
• At most k zeros in the window
• Shrink while the window is invalid
• Longest valid window expands greedily

All sliding window lessons

Other patterns

• Hash map
• Hash set
• Array state
• Two pointers