The AWS Certified Machine Learning Engineer - Associate is changing version this autumn. If you are studying now, know this: the current exam, MLA-C01, has its last English sitting on 28 September 2026, and registration for the updated MLA-C02 opens on 1 September 2026. Below is exactly what that means and how to decide whether to sit the old exam or wait.

What the certification is

This associate-level credential validates that you can build, deploy and operate machine learning workloads in production on AWS, not just train a model in a notebook. AWS aims it at people with at least one year of experience using Amazon SageMaker and other ML engineering services, and lists target roles including backend developer, DevOps engineer, data engineer, MLOps engineer and data scientist. It sits a step beyond the more conceptual AWS Certified AI Practitioner, which is a good starting point if this one feels too deep.

The exam format

For the current MLA-C01, taken from AWS:

  • 65 questions
  • 130 minutes
  • 150 US dollars
  • Delivered at a Pearson VUE test centre or as an online proctored exam
  • Available in English, Japanese, Korean and Simplified Chinese

The exam covers four areas: Data Preparation for Machine Learning, ML Model Development, Deployment and Orchestration of ML Workflows, and ML Solution Monitoring, Maintenance and Security. Download the official exam guide from AWS for the current domain weightings and task statements, and study against those rather than any third-party summary.

The version change, and the dates that matter

AWS has confirmed the transition on the certification page. The key milestones:

  • 1 September 2026: registration opens for the updated exam, MLA-C02.
  • 28 September 2026: the last day to take the current MLA-C01 in English.
  • The current exam remains available in Korean, Japanese and Simplified Chinese until MLA-C02 reaches general availability.

Version updates like this refresh the services, question style and emphasis to match how AWS expects the work to be done today. Incumbent study guides and courses written for MLA-C01 go stale the moment C02 lands, so be careful what you buy in the changeover window. For the official outline and any new preparation resources, check the AWS certification page directly on or after 1 September.

Should you sit C01 now or wait for C02?

Sit MLA-C01 before 28 September if you are already well into preparation in English and ready to test. A pass is a pass; your certification does not lose value because a new version exists, and AWS certifications are valid for three years.

Wait for MLA-C02 if you are only starting out, if you sit in one of the languages that keeps C01 longer and would rather learn the current outline, or if you want your study material to match the exam exactly. Rushing an unfinished C01 attempt just to beat the deadline is a false economy.

How to prepare either way

Ground your study in the official exam guide and in hands-on SageMaker work, because this exam rewards real pipeline experience over memorised facts. Build, deploy and monitor a small end-to-end model so the deployment and monitoring domains are muscle memory, not theory. If you are still choosing your first cloud credential, our note on the Cloud Digital Leader timing shows how much these deadline windows can matter.

Practise under real exam conditions

The MLA exam is scenario-heavy, and 65 questions in 130 minutes leaves no room to freeze. Full, timed mock exams train the pacing and the question style, not just the content. On CandidatesPrep you can sit complete simulations scored by domain, so you can see whether it is model development or deployment and monitoring that is costing you, and repeat until every area is solid. Trainers moving a cohort through the C01-to-C02 change can track readiness centrally and steer everyone onto the right version.

The bottom line

The AWS Machine Learning Engineer - Associate is mid-transition. If you are ready and sit in English, take MLA-C01 before 28 September 2026. Otherwise, prepare for MLA-C02 from 1 September and study against the current official guide. Either way, let the exam version, not the panic of a deadline, decide your date.

Want to know if you are exam-ready? Sit a full timed simulation, or book a demo if you train cloud and ML candidates through version changes like this.