Amazon Coding Assessment Practice Test PDF 2026

Amazon Coding Assessment Practice Test PDF — Free Study Guide for SDE Interviews

Preparing for Amazon's software development engineering (SDE) hiring process requires a specific approach that balances algorithmic problem-solving with behavioral competency. A Amazon coding assessment practice test PDF gives you printable problems, solution patterns, and concept reviews to work through offline. Download our free PDF below and use it alongside your LeetCode practice and Amazon Leadership Principles preparation to build a complete interview readiness strategy.

Amazon SDE Hiring Process Overview

Amazon's engineering hiring pipeline is structured, multi-stage, and designed to assess both technical depth and cultural alignment with the company's Leadership Principles. Understanding the full pipeline helps you allocate preparation time effectively rather than over-preparing for one stage at the expense of another.

The typical SDE hiring process begins with an online assessment (OA) — the coding assessment this guide focuses on. Candidates who pass the OA are invited to a phone screen (usually one technical interview with a recruiter or engineer). Those who advance from the phone screen proceed to a full onsite loop — historically conducted in person at an Amazon office, now frequently conducted virtually. The onsite loop typically consists of four to six rounds including one or two dedicated "Bar Raiser" interviews led by senior engineers or managers from outside the hiring team.

Each round in the onsite loop tests a combination of coding, system design, and behavioral questions. The balance shifts by level: SDE-I roles emphasize coding and data structures heavily. SDE-II roles introduce more system design. Senior SDE and principal-level roles are primarily system design and behavioral, with coding as a secondary signal. This guide focuses on the online assessment and coding components — but behavioral preparation runs in parallel and is equally critical to the offer decision.

Online Assessment Format

The Amazon SDE online assessment is typically administered through HackerRank or a similar platform. The standard format presents two algorithmic coding problems to be solved within 70 minutes. You write and run code in your chosen language — Python, Java, C++, JavaScript, and others are typically available. The platform evaluates your solutions against a set of hidden test cases, including edge cases.

The two problems are usually of mixed difficulty: one problem at LeetCode Medium difficulty level and one at Medium-Hard. Occasionally both are Medium, or the second problem ramps to Hard. The combination is designed so that a strong candidate can complete both correctly within the time limit, while weaker candidates struggle with the second. Partial credit exists — your solution scores points based on how many test cases it passes, so a correct-but-incomplete solution on problem two is better than abandoning it entirely.

Beyond the two coding problems, some Amazon OA formats include a work simulation component or a work style assessment that tests personality and behavioral alignment with Leadership Principles. These sections are not code-based — they use scenario-response questions and cannot be "practiced" in the traditional sense, though familiarity with the Leadership Principles helps you respond authentically and consistently.

Time management matters enormously. With 70 minutes for two problems, you have roughly 35 minutes per problem. If you spend 45 minutes on problem one without a working solution, you leave insufficient time for problem two. A strategy of getting a working-but-inefficient solution first, then optimizing, is usually better than pursuing the optimal solution from the start and running out of time.

Core Data Structures You Must Know

Amazon coding problems draw from a well-defined set of data structures that appear repeatedly across the LeetCode-style problem bank. Solid command of these structures — knowing their operations, time complexities, and typical use cases — is non-negotiable for SDE interview readiness.

Arrays and Strings: The most fundamental structure. Problems involving subarrays, substrings, rotation, reversal, and in-place modification are extremely common. You must be able to traverse arrays in multiple directions, use index arithmetic fluently, and handle edge cases involving empty arrays and single-element arrays without bugs.

Hash Maps (Dictionaries): Hash maps provide O(1) average lookup, insert, and delete. They are the most powerful tool for reducing time complexity — converting O(n²) brute force solutions to O(n) with space trade-off. Problems involving frequency counting, anagram detection, two-sum variants, and index tracking all benefit from hash map solutions. You should be able to implement hash map solutions fluently in any language you use.

Stacks and Queues: Stacks support last-in-first-out (LIFO) operations; queues support first-in-first-out (FIFO). Stacks appear in bracket matching, expression evaluation, monotonic stack problems, and DFS implementations. Queues appear in BFS, level-order tree traversal, and sliding window problems. Deques (double-ended queues) generalize both and appear in sliding window maximum problems.

Trees and Binary Trees: Binary tree problems are a mainstay of Amazon interviews. You must implement recursive and iterative traversals (inorder, preorder, postorder), work with binary search trees (BST insert/delete/search), and understand tree balancing concepts. Common problem types include finding lowest common ancestor, validating BSTs, path sum problems, and level-order traversal.

Graphs: Graph problems appear frequently at the SDE-II level and above. You need working implementations of depth-first search (DFS) and breadth-first search (BFS) for both adjacency list and matrix representations. Connected components, cycle detection, shortest path (Dijkstra for weighted, BFS for unweighted), and topological sort are recurring patterns. Graph problems often disguise themselves as matrix traversal problems — recognizing the connection helps.

Linked Lists: Classic linked list problems include reversal (iterative and recursive), cycle detection (Floyd's tortoise and hare), merging sorted lists, finding the middle node, and removing duplicates. Two-pointer techniques appear throughout. You should be comfortable with both singly linked and doubly linked list implementations.

Algorithm Paradigms and Problem Patterns

Recognizing which algorithmic paradigm applies to a problem is the most important skill separating candidates who solve problems quickly from those who struggle. These patterns repeat across hundreds of problems in slightly different forms — once you internalize them, you can identify the right approach within the first 30 seconds of reading a problem.

Two Pointers: A pair of indices traverses the data structure simultaneously, reducing O(n²) nested loop solutions to O(n). Classic applications: two-sum on sorted array, valid palindrome check, remove duplicates from sorted array, container with most water. The key insight is that you can use pointer movement to efficiently explore the solution space without examining every pair.

Sliding Window: A variable-size or fixed-size window slides through an array or string, maintaining state about the window's contents. Applications: maximum sum subarray, longest substring without repeating characters, minimum window substring. The critical implementation detail is handling window expansion and contraction correctly at the boundaries.

Depth-First Search (DFS): Recursive or stack-based exploration that goes as deep as possible before backtracking. Applications: tree traversal, path finding, island counting in matrix problems, combinations/permutations generation. The recursion pattern for DFS is extremely consistent — learn it once and apply it broadly.

Breadth-First Search (BFS): Queue-based level-by-level exploration. Applications: shortest path in unweighted graphs, level-order tree traversal, minimum steps problems. BFS guarantees finding the shortest path in unweighted graphs — DFS does not. Choosing between them depends on whether shortest path matters.

Dynamic Programming: Optimization over recursive subproblems with overlapping substructure. DP is the most feared topic among candidates but also the most learnable — the same pattern (define state, write recurrence, fill table or memoize) applies to dozens of problems. Key DP patterns: 1D DP (climbing stairs, house robber), 2D DP (grid paths, edit distance, coin change), interval DP, and knapsack variants. Start with the simplest patterns and build up systematically.

Binary Search: Not just for sorted arrays — binary search applies to any monotone function over a range. Classic applications: find first/last position, rotated sorted array search, search in matrix, and finding the minimum in a range that satisfies a condition. The template is standard; the creative application requires recognizing when the problem has a monotone structure to exploit.

Amazon Leadership Principles in Coding Interviews

Unlike most tech company interviews, Amazon explicitly evaluates Leadership Principles (LPs) in every interview round — including technical rounds. Interviewers at Amazon are trained to assess LP alignment throughout the conversation, not just in dedicated behavioral sections. This means your coding interview can be partly a behavioral evaluation based on how you handle uncertainty, ask clarifying questions, discuss trade-offs, and respond to hints.

The 16 Leadership Principles include Customer Obsession, Ownership, Invent and Simplify, Are Right A Lot, Learn and Be Curious, Hire and Develop the Best, Insist on the Highest Standards, Think Big, Bias for Action, Frugality, Earn Trust, Dive Deep, Have Backbone; Disagree and Commit, Deliver Results, Strive to be Earth's Best Employer, and Success and Scale Bring Broad Responsibility.

For coding interviews, the most relevant principles in practice are: Dive Deep (showing detailed technical knowledge rather than surface-level answers), Bias for Action (proceeding with a reasonable approach rather than over-analyzing), Invent and Simplify (finding elegant solutions rather than brute force patches), and Deliver Results (getting to working code even when the optimal solution isn't immediately clear).

Behavioral Interview: STAR Method

Amazon's behavioral interviews use the STAR method: Situation, Task, Action, Result. Every behavioral response should follow this structure precisely. Amazon interviewers specifically probe for your individual contribution ("what did you do" versus "what did the team do"), the measurable impact of your actions, and the scale of what you were working on.

Prepare 8–10 strong behavioral stories before your interview. Each story should be specific, metric-driven, and clearly articulate your individual decision-making and actions. Interviewers will probe with follow-up questions — "what would you have done differently," "what was the biggest obstacle," "how did you measure success." Stories that can sustain this depth of follow-up are significantly more compelling than surface-level anecdotes.

Cover at minimum: a time you failed and what you learned, a time you disagreed with your manager or team and what happened, a time you took ownership of a problem outside your scope, a time you had to make a decision with incomplete data, and a time you significantly improved a process or system. These are the most commonly asked LP story types across all Amazon interview levels.

System Design Basics for SDE II Roles

SDE-II and above candidates face dedicated system design rounds in the onsite loop. System design interviews assess your ability to architect scalable, reliable, distributed systems under constraints — a fundamentally different skill from algorithm problem solving.

Core system design concepts you need to know: horizontal vs. vertical scaling, load balancing (round-robin, least connections, consistent hashing), caching strategies (CDN, application cache, database cache, eviction policies like LRU), database choices (relational vs. NoSQL, sharding, replication, eventual consistency vs. strong consistency), message queues and event-driven architecture, microservices vs. monolithic design, and API design (REST vs. GraphQL vs. gRPC).

A well-structured system design answer follows a consistent framework: clarify requirements and scale expectations, estimate traffic and storage, define the API/interface, design the high-level architecture, deep-dive into the most complex components, discuss trade-offs explicitly, and address failure modes. Interviewers reward candidates who discuss trade-offs proactively rather than defending a single design as optimal in all scenarios.