Databricks Data Engineer Associate exam: what's on it and how to prepare
The Databricks Certified Data Engineer Associate exam in one page: format, cost, the seven domains and their weights, the product names that changed, and a step-by-step plan to get ready.
By Alex Cole, Databricks MVP and Solutions Architect Champion · · Facts checked against the official docs on
The Databricks Certified Data Engineer Associate exam checks that you can do foundational data engineering on the Databricks Platform: ingest data, transform it with PySpark and SQL, orchestrate it with Lakeflow Jobs, ship it with CI/CD, keep it running, and govern it with Unity Catalog. This post covers the current exam (published 4 May 2026, guide updated 6 August 2026): the facts, what each domain asks of you, and how to prepare.
The exam at a glance
| Scored questions | 45, multiple choice |
| Unscored questions | Up to 10 more, not marked as such; extra time is already included |
| Time limit | 90 minutes |
| Cost | USD 200, plus local tax |
| Delivery | Online proctored or at a test centre |
| Exam languages | English, Japanese, Brazilian Portuguese, Korean |
| Data manipulation code | SQL where possible, otherwise Python |
| Test aids | None allowed |
| Prerequisites | None; Databricks highly recommends the related training |
| Recommended experience | Six months of hands-on work with the tasks in the exam outline (from the exam guide) |
| Valid for | Two years |
What the exam covers
The exam has seven domains. Databricks publishes the weights on the certification page:
| Domain | Weight | What you need to be able to do |
|---|---|---|
| Databricks Intelligence Platform | 6% | Explain the core components (Delta Lake, Unity Catalog) and pick the right compute for a workload |
| Data Ingestion and Loading | 21% | Load batch, streaming and incremental data with COPY INTO, Auto Loader, Lakeflow Connect and JDBC/REST clients, including nested JSON, and choose a method by volume, frequency, data type and governance needs |
| Data Transformation and Modeling | 22% | Clean bronze data into silver tables, join, reshape, deduplicate and aggregate DataFrames, tune basic Spark settings, build gold objects such as materialized views and streaming tables, and apply data quality checks to silver and gold data |
| Working with Lakeflow Jobs | 16% | Build multi-task jobs (notebook, SQL query, dashboard and pipeline tasks) with dependencies, retries, branching and looping, and schedule them with time-based or data-driven triggers |
| Implementing CI/CD | 10% | Work with Git folders, and package and promote code across dev, test and prod with Declarative Automation Bundles (including per-target variables and overrides) and the Databricks CLI |
| Troubleshooting, Monitoring, and Optimization | 10% | Read job run history and the Spark UI, spot skew, shuffle and spill, know Liquid Clustering and predictive optimization, and diagnose cluster startup failures, library conflicts and out-of-memory errors |
| Governance and Security | 15% | Manage managed and external tables, grant, revoke and deny privileges, and apply row filters, column masks and ABAC policies |
Transformation and ingestion together are 43% of the exam, so most of your practice time belongs there.
Names that changed
Several products on this exam have new names. Many courses and older practice material still use the old ones:
| Older name | Current name | Notes |
|---|---|---|
| Databricks Asset Bundles | Declarative Automation Bundles | The exam guide uses the new name |
| Databricks Repos | Git folders | The exam guide uses the new name |
| Databricks Workflows | Lakeflow Jobs | Earlier docs and courses say Workflows; the exam guide says Lakeflow Jobs |
| Delta Live Tables (DLT) | Lakeflow pipelines | The exam guide calls them Lakeflow Declarative Pipelines |
How to prepare
- Read the official exam guide on the certification page, and work through its sample questions. For five more with the reasoning shown step by step, see Data Engineer Associate practice questions: worked examples.
- Take the free diagnostic. It samples every domain in about 10 minutes, with no sign-up, and shows where you'd lose marks.
- Do the recommended training. The exam guide lists the instructor-led course Data Engineering with Databricks, and these self-paced courses in Databricks Academy: Data Ingestion with Lakeflow Connect, Deploy Workloads with Lakeflow Jobs, DevOps Essentials for Data Engineering, Build Data Pipelines with Apache Spark Declarative Pipelines and Getting started with Governance. Start from the Databricks Academy login page, which sends you to the right academy: Customer Academy for customers and anyone learning on their own, Partner Academy for people at consulting and software partners, and Microsoft Academy for Microsoft employees.
- Get hands-on. The exam guide points to Databricks Free Edition for small practice tasks. It's free for personal learning, with smaller compute than a paid workspace. For example: load files with Auto Loader, build a two-task job, deploy a bundle, and grant access to a table. CertLake's PySpark koans are short coding exercises that run in your browser, and some are free with a free account.
- Study the weak domains. Use the study guide for the concepts and the flashcards to make them stick, starting with the domains the diagnostic flagged.
- Book the exam once your practice scores are steady, and add the date in Settings so your CertLake dashboard counts down to it.
- In the final week, take a timed practice exam in one sitting and revise from the cheat sheet. Re-read the official exam guide: Databricks asks you to check it two weeks before your exam in case it has changed.
On exam day
- For an online exam, run the proctoring system check well before the day, not on the morning.
- No test aids are allowed.
- Unscored questions aren't marked, so give every question the same effort.
Frequently asked questions
How many questions are on the Databricks Data Engineer Associate exam?
45 scored multiple-choice questions, plus up to 10 unscored questions that look the same. You have 90 minutes in total, which already allows for the unscored ones.
How much does the Data Engineer Associate exam cost?
USD 200, plus any tax due where you live.
What is the pass mark?
Databricks doesn't state a pass mark in the exam guide or on the certification page. CertLake's practice exams use 70% as the pass line, so aim to score comfortably above that in timed practice.
Is the exam in Python or SQL?
Both. Databricks says data manipulation code in the exam is shown in SQL where possible, and in Python otherwise.
Are there any prerequisites?
No. Databricks recommends the related training and about six months of hands-on experience with the tasks in the exam outline.
How long is the certification valid?
Two years. To stay certified, you recertify every two years. The exam guide says you can take the Data Engineer Associate renewal exam or the current full exam; the certification page mentions only the current exam, so check which applies before you book.
See where you stand on Data Engineer Associate
The free diagnostic shows which exam domains you'd lose marks on, in about 10 minutes and without signing up.
Spotted something out of date? Email support@certlake.com.