Skip to content

AWSMLA

MLA Practice Exam — AWS Certified Machine Learning Engineer - Associate

MLA is an engineering exam wearing an ML hat: most questions are about preparing data, shipping a model to an endpoint, and keeping it healthy, rather than deriving algorithms. Practise across all four official domains with full 65-question papers and a score breakdown per domain.

MLA exam at a glance

Questions per paper
65
ReadyForCert Originals
67 questions
Question bank
225 questions
Time limit
130 minutes
Passing score
720 / 1000
Question format
Multiple choice and multiple response
Official exam fee
USD $150
Certification valid for
3 years
ReadyForCert access
60 days · $8.99 AUD
Free sample
20 questions

Study, practise, review

Study by topic

Work through the complete question bank by official exam domain, read the explanation after each answer, save questions, and revisit incorrect ones.

Timed exams

Take 65-question simulations with a 130-minute timer and saved progress.

Exam reports

Review your score, every answer with its explanation, and your performance across each exam domain.

What the MLA exam covers

  1. Data Preparation for ML and AI

    Collecting and storing data, transforming it and engineering features, and checking data quality and bias.

    28%
  2. ML Model and Foundation Model (FM) Development

    Choosing a modelling approach, training, fine-tuning and customising models, and evaluating how they perform.

    24%
  3. Deployment and Orchestration of ML and AI Workflows

    Deployment infrastructure for each model type, provisioning resources, and CI/CD pipelines for MLOps.

    24%
  4. Operating, Monitoring, and Securing ML and AI Solutions

    Monitoring inference and model performance, managing cost and capacity, and securing workloads and endpoints.

    24%

A sample MLA question

One question written for this page, in the same style as the questions in the bank.

Operating, Monitoring, and Securing ML and AI Solutions

A fraud-detection model deployed to an Amazon SageMaker AI real-time endpoint keeps its accuracy on the held-out test set, but its precision in production falls steadily over three months. Which action identifies the cause?

  • ARetrain the model every week on the most recent month of labelled transactions and redeploy it.
  • BMove the endpoint to a larger instance type and turn on auto scaling to cut inference latency.
  • CSet Amazon CloudWatch alarms on the endpoint's invocation errors and model latency.
  • DSchedule drift detection that compares recent inference inputs with the training baseline and alerts on drift. — Correct answer

About the MLA exam

How many questions are on the MLA exam?

65 questions in 130 minutes. Only 50 of them affect your score; the other 15 are unscored questions AWS uses to trial new items. You cannot tell which is which, so treat all 65 as real. AWS is updating this exam. In English, only the updated beta is on offer for now: 85 questions in 170 minutes, USD $75 at beta pricing. ReadyForCert papers keep the 65-question, 130-minute format until the updated exam is generally available.

What score do I need to pass MLA?

AWS reports a scaled score from 100 to 1,000 and the minimum passing score is 720. The scale is not a straight percentage of questions answered correctly, so ReadyForCert reports your score against 87%: a higher bar that leaves you a margin on the day.

Do I need to be a data scientist to pass MLA?

No. MLA targets engineers who put ML into production. It expects you to be comfortable with pipelines, endpoints, CI/CD and monitoring; it does not expect you to derive algorithms or tune models the way the specialty-level exam did.

Are these the real MLA exam questions?

No, and you should be sceptical of anyone who claims otherwise. These are original questions written against the current AWS exam guide. Real exam content is under NDA, and AWS rotates its bank — memorising a dump is the fastest way to fail a refreshed exam.