The Most-Asked Amazon Interview Questions in 2026 (From 4,577 Reports)
We pulled every Amazon question in the LeakCode database and ran the numbers. 4,577 clean entries. Here is what the data shows about Amazon's coding rounds, OAs, and system design in 2026.
Amazon has 8,133 interview reports in the LeakCode database as of May 2026 — the highest of any company we track. After filtering for duplicates and junk, 4,577 distinct questions remain. This post is an analysis of those 4,577 entries: where they came from, what types of questions they represent, and which topics appear most often in Amazon's loops.
All numbers below are real counts pulled directly from the LeakCode database. Nothing is estimated or extrapolated.
Where the data comes from
The 4,577 Amazon questions come from seven sources. LeetCode Discuss dominates because Amazon's interview volume is large and English-speaking candidates report there most actively. The 1Point3Acres OJ catalog (715 entries) adds signal from Chinese-language reporting that most Western prep resources miss entirely.
| Source | Questions | Share |
|---|---|---|
| LeetCode Discuss | 2,751 | 60.1% |
| 1p3a OJ Catalog | 715 | 15.6% |
| LeetCode Company Lists | 594 | 13.0% |
| 256 | 5.6% | |
| GeeksforGeeks | 141 | 3.1% |
| 1Point3Acres Forum | 83 | 1.8% |
| InterviewDB | 31 | 0.7% |
| Blind | 6 | 0.1% |
The split matters for understanding coverage. LeetCode Discuss entries tend to be full interview experience posts with round-by-round narrative. The LeetCode company list entries are structured mappings of specific problems to Amazon. The 1p3a OJ catalog entries are crowd-sourced reports of which LeetCode problem IDs appeared in Amazon's coding rounds. These three data types are complementary, and each fills gaps the others miss.
Round type breakdown
Amazon's interview process typically involves an online assessment, one or more phone screens, and a virtual onsite. Behavioral rounds (Amazon calls them "bar raiser" interviews) run parallel to the technical loops. The LeakCode data reflects this structure clearly.
| Round Type | Questions | Share |
|---|---|---|
| Online Assessment (OA) | 1,415 | 30.9% |
| Coding | 1,302 | 28.4% |
| Phone Screen | 550 | 12.0% |
| System Design | 269 | 5.9% |
| Behavioral | 254 | 5.5% |
| Onsite | 143 | 3.1% |
| Recruiter Screen | 136 | 3.0% |
| Other / Unclassified | 508 | 11.1% |
OAs are the plurality. Amazon runs a mandatory online assessment for almost all SDE roles before scheduling a phone screen. The OA typically involves two coding problems on HackerRank, timed at 90 minutes. The coding and phone screen rows cover the synchronous rounds that follow the OA for candidates who pass.
The 254 behavioral entries represent the leadership principles component. Amazon's bar raiser process means behavioral rounds are evaluated separately from technical performance, and a strong coder who struggles on LP questions can still be rejected. Prep for both.
Top topics by frequency
We classify every question in the LeakCode database by topic using automated tagging. Here is how Amazon's question set distributes across topics.
| Topic | Questions |
|---|---|
| System Design | 516 |
| Arrays | 456 |
| Graph | 291 |
| Strings | 272 |
| Dynamic Programming | 236 |
| Behavioral / Leadership Principles | 176 |
| Binary Tree | 141 |
| Hash Table | 133 |
| Greedy | 113 |
| Heap / Priority Queue | 98 |
| Sliding Window | 93 |
| Binary Search | 70 |
System design is the top topic by count, which reflects that senior roles (SDE2 and above) always include a dedicated system design round. For new grad and SDE1 roles, system design is less common but still appears occasionally. Arrays, graphs, and strings together account for over 1,000 questions, confirming that Amazon's coding rounds lean toward classic data structure problems rather than exotic algorithmic tricks.
The most common coding problems
The LeetCode company list source gives us the clearest signal on specific problems. These are crowd-sourced reports of which LeetCode problems appeared in Amazon's coding rounds, tagged by number. The list below is sorted by problem number; these are all problems that appear in the Amazon tag in the LeakCode database from this source alone.
A selection of the highest-frequency problems from Amazon's coding rounds, based on cross-source appearance:
| Problem | Topic |
|---|---|
| #1 Two Sum | Hash Table |
| #2 Add Two Numbers | Linked List |
| #3 Longest Substring Without Repeating Characters | Sliding Window |
| #15 3Sum | Two Pointers |
| #20 Valid Parentheses | Stack |
| #21 Merge Two Sorted Lists | Linked List |
| #23 Merge k Sorted Lists | Heap |
| #33 Search in Rotated Sorted Array | Binary Search |
| #42 Trapping Rain Water | Two Pointers |
| #49 Group Anagrams | Hash Table |
| #53 Maximum Subarray | Dynamic Programming |
| #78 Subsets | Backtracking |
| #146 LRU Cache | Hash Table / Linked List |
| #200 Number of Islands | Graph / BFS |
| #207 Course Schedule | Graph / Topological Sort |
| #297 Serialize and Deserialize Binary Tree | Binary Tree |
| #322 Coin Change | Dynamic Programming |
| #347 Top K Frequent Elements | Heap |
| #560 Subarray Sum Equals K | Hash Table / Prefix Sum |
| #886 Possible Bipartition | Graph |
| #907 Sum of Subarray Minimums | Stack |
| #992 Subarrays with K Different Integers | Sliding Window |
The full list of problems in the LeakCode database for Amazon is accessible at the Amazon company page. The table above is a representative sample of the highest-signal entries.
What Amazon's OAs actually look like
The 1,415 OA-tagged entries in our database come from candidates who reported their specific OA questions. Amazon's OAs are delivered via HackerRank and vary by role and region, but common patterns emerge from the reports.
Most OAs contain two algorithmic coding problems to be solved in 90 minutes. The difficulty range in the reports runs from medium to hard (LeetCode scale). Problems frequently involve arrays with a custom constraint, simulation problems with an inventory or resource management theme, or graph traversal on a grid. The simulation problems in particular (inventory migration, warehouse routing, item placement) appear repeatedly and reflect Amazon's logistics-heavy product surface.
There are also OA reports describing a work simulation component, separate from the coding questions, used for behavioral pre-screening in some hiring tracks. LeakCode does not index these as coding questions; they appear in the behavioral category.
System design at Amazon
System design rounds appear for SDE2 and above, and increasingly for strong SDE1 candidates in competitive locations. The 516 system design-tagged questions in the Amazon data include reports of:
- Design a URL shortener
- Design Amazon's order management system
- Design a distributed cache
- Design a notification service at scale
- Design the Amazon locker system (an Amazon-specific domain question)
- Design a recommendation engine
The locker design problem appears in multiple independent reports, making it one of the more reliably documented Amazon-specific system design questions in the database.
How to use this data
The goal of this analysis is to surface the shape of Amazon's interview process, not to give you a magic list to memorize. LeetCode problems rotate. OA questions change between cohorts. What persists across reports is the topic distribution: arrays, graphs, strings, DP, trees, system design at senior levels.
Use the topic table to weight your prep. If you are targeting SDE1, prioritize arrays, strings, graphs, and trees. If you are targeting SDE2, add system design at the same priority level. In both cases, plan dedicated time for behavioral prep using the Amazon Leadership Principles framework.
Browse the full Amazon question set on the Amazon company page. The LeakCode stats page shows the current total count and last update date.