The Databricks Certified Data Analyst Associate exam validates that you can use Databricks SQL, AI/BI dashboards and Genie spaces to find, query, analyze, visualize and secure data on the lakehouse. It is a 45-question, 90-minute proctored exam that costs US$200, spans nine sections and 39 objectives, and has no published passing score.
Last updated September 2026.
The exam is 45 scored multiple-choice questions in 90 minutes, delivered online with a proctor or at a test center. Some items ask you to select two answers. There is no prerequisite; Databricks recommends related course attendance and about six months of hands-on experience as a data analyst. Registration is US$200, the certification is valid for two years, and recertifying means taking the full live exam again. The form may include unscored questions that are not identified and do not affect the score, with extra time factored in.
The official guide, dated 30 October 2025, organizes the material into nine sections covering 39 objectives. Databricks publishes neither per-section percentages nor a cut score, so the objective counts below are the only honest guide to emphasis.
| Section | Objectives | Share of objectives |
|---|---|---|
| 1. Understanding the Data Intelligence Platform | 3 | ~8% |
| 2. Managing Data | 3 | ~8% |
| 3. Importing Data | 2 | ~5% |
| 4. Executing Queries with Databricks SQL and SQL Warehouses | 9 | ~23% |
| 5. Analyzing Queries | 6 | ~15% |
| 6. Working with Dashboards and Visualizations | 7 | ~18% |
| 7. Developing, Sharing, and Maintaining AI/BI Genie Spaces | 4 | ~10% |
| 8. Data Modeling with Databricks SQL | 2 | ~5% |
| 9. Securing Data | 3 | ~8% |
Objective counts are transcribed from the official Databricks Data Analyst Associate exam guide (30 October 2025 edition). The share column is derived from those counts and is shown only to indicate relative emphasis; it is not an official weighting.
Sections 4, 5 and 6 together hold 22 of the 39 objectives: writing SQL on a warehouse, reading why a query is slow, and turning results into dashboards and alerts. That is the working core of the job, and the exam treats it that way. Everything else is the governance and product surface around it.
Choosing a SQL warehouse (serverless is the recommendation), querying external databases in place through Lakehouse Federation, and knowing when a streaming table, a materialized view or a dynamic view is the right object. Then the SQL itself: aggregates and HAVING, join types and the row explosion a bad key causes, UNION versus UNION ALL, sorting with nulls, CREATE TABLE AS SELECT, and Delta time travel with VERSION AS OF and RESTORE.
What Photon speeds up and what it does not, finding a slow statement in Query History and reading its query profile, auditing with DESCRIBE HISTORY, how the result cache decides to serve a stored answer, and Liquid Clustering as the recommended table layout. Partitioning and Z-Order appear only as the contrast case.
AI/BI dashboards: datasets, widgets, parameters, publishing with shared or individual data permissions, schedules, subscriptions and SQL alerts. Genie spaces: what a space is built from (datasets, a knowledge store, instructions, trusted assets), the permission levels, embedding, and the monitor-feedback-benchmark loop that keeps one accurate. Expect the credential model to be tested: queries run on the author's warehouse credentials, data access is always the asking user's.
The Unity Catalog object model and three-level namespace, certified tables and lineage in Catalog Explorer, Marketplace and the ingestion paths, star versus snowflake versus data vault and where each sits in the medallion architecture, and securing personal data with ownership, column masks, row filters, dynamic views and tags. Primary and foreign key constraints are informational, never enforced; that one fact shows up more than once.
The exam guide is from October 2025, and Databricks renamed several of its products during 2026. The guide's names are the ones the questions use, so learn both:
| Exam guide says | 2026 docs say | Since |
|---|---|---|
| Databricks Assistant | Genie Code | 11 March 2026 |
| AI/BI Genie space | Genie Agent | July 2026 |
| Delta Sharing | OpenSharing | 2026 documentation |
| Delta Live Tables (older material) | Lakeflow Spark Declarative Pipelines | 2025 |
Rename dates are from the Databricks platform and AI/BI release notes. The capabilities did not change with the names.
For a week-by-week plan by background, see how long it takes to prepare for the Data Analyst Associate exam.
The exam has 45 scored multiple-choice questions with a 90-minute time limit, and some items ask you to select two answers. Unscored questions can also appear without being identified, with extra time factored in, so the form may show more than 45.
Databricks does not publish a passing score for this exam. Aim to score consistently well on full-length practice exams rather than chasing a specific percentage.
Nine sections: the Data Intelligence Platform, managing data, importing data, executing queries with Databricks SQL, analyzing queries, dashboards and visualizations, Genie spaces, data modeling, and securing data. Databricks SQL, query analysis and dashboards hold 22 of the 39 objectives between them.
No prerequisite is required. Databricks recommends related course attendance and about six months of hands-on experience as a data analyst, and the scenario questions assume you have used Databricks SQL and AI/BI dashboards.
Registration is US$200 and the exam is delivered online or at a test center with a proctor. The certification is valid for two years, after which you take the full live exam again.
Certify's Data Analyst Associate course covers all nine sections across 43 mobile-first lessons, with four section exams and three full-length mocks. Every chapter is free.
Start the course →