Microsoft · DP-600

DP-600 Exam: Complete Microsoft Fabric Analytics Guide

Current DP-600 exam guide covering the $165 USD price, 100-minute limit, skills measured, preparation priorities, semantic models, and renewal.

Last verified September 6, 2026

DP-600 exam at a glance

Facts last verified September 6, 2026

Domain breakdown

Who should take the DP-600 exam?

DP-600 is for analytics professionals who design, create, and manage enterprise analytical assets in Microsoft Fabric. Typical candidates are experienced Power BI developers, analytics engineers, data analysts with platform responsibilities, and data engineers who own the semantic and consumption layers of a Fabric solution.

Microsoft expects you to prepare and enrich data, secure and maintain analytics assets, and implement semantic models. You should be able to query and analyze data with SQL, KQL, and DAX and work with stakeholders, architects, engineers, analysts, and administrators. This is an intermediate exam, not a general introduction to Fabric.

There is no required certification before DP-600. In practice, candidates should understand star schemas, relationships, filter context, data quality, access control, and application lifecycle management. If you have only built simple Power BI reports, add hands-on work with warehouses, lakehouses, OneLake, Eventhouses, Git integration, and deployment pipelines.

Skills measured on DP-600

The official DP-600 study guide lists the skills measured as of July 21, 2026. Data preparation is the largest domain at 45–50%, while solution maintenance and semantic models each account for 25–30%.

Maintain a data analytics solution (25–30%)

Security and governance span workspace- and item-level access plus row-, column-, object-, and file-level controls. Know where each boundary is enforced and how sensitivity labels and endorsement communicate trust and handling requirements. Scenario questions often ask for the narrowest control that satisfies a business requirement.

The development lifecycle includes workspace version control, Power BI Desktop project files, and Fabric deployment pipelines. You must understand how content moves through development, test, and production and how to assess downstream impact across lakehouses, warehouses, dataflows, and semantic models before making a change.

Microsoft also names XMLA endpoint management and reusable assets such as PBIT templates, PBIDS data-source files, and shared semantic models. Practice deploying and managing a model programmatically and deciding when a reusable shared model is better than duplicated logic in separate reports.

Prepare data (45–50%)

This domain starts with data access. Be able to create connections, discover assets through the OneLake catalog and Real-Time hub, choose a store, and decide whether data should be ingested or accessed in place. Understand OneLake integration for Eventhouses and semantic models rather than treating every workload as a separate copy of the data.

Transformation objectives include views, functions, stored procedures, calculated columns or tables, star schemas for lakehouses and warehouses, denormalization, aggregation, joins, and data-type conversions. You must also identify and resolve duplicates, nulls, missing records, and other quality issues. Build repeatable transformations instead of fixing a sample dataset manually.

DP-600 tests selection, filtering, and aggregation through the Visual Query Editor, SQL, KQL, and DAX. Learn the purpose of each language and practice equivalent basic operations across them. The goal is not identical syntax; it is knowing which engine and layer should perform a transformation or analytical query.

Implement and manage semantic models (25–30%)

Model design includes choosing a storage mode, building a star schema, and implementing relationships such as bridge tables and many-to-many patterns. You should write DAX with variables, iterators, table filters, window functions, and information functions, then explain the effect of row and filter context.

Enterprise models also use calculation groups, dynamic format strings, field parameters, large-model storage, and composite designs. Do not study these as isolated features. Build a model where each feature removes duplication, improves usability, or satisfies a scale requirement.

Optimization covers DAX, queries, and report visuals. Know how to configure Direct Lake fallback and refresh behavior, choose between Direct Lake on OneLake and on the SQL analytics endpoint, and implement incremental refresh. Measure before tuning and connect a symptom to the model, storage mode, query plan, visual, or capacity behavior responsible for it.

How to prepare for DP-600

Build one coherent analytics solution. Load data into a lakehouse or warehouse, clean and model it, expose a governed semantic model, write DAX measures, connect a report, configure security, and deploy through source control. Then introduce a breaking schema change and use impact analysis to understand the consequences.

Study decision boundaries: warehouse versus lakehouse, import versus Direct Lake versus DirectQuery, SQL versus KQL versus DAX, and shared model versus report-specific model. Keep short “use when / avoid when” notes and validate them in Fabric rather than memorizing only definitions.

A six-week plan can cover Fabric and OneLake foundations, data preparation, semantic design, advanced DAX, lifecycle and security, then performance and timed review. Spend close to half your study time on data preparation because it has the largest official weight.

DP-600 practice questions

Use the free Practice Assessment from the official certification page to find weak objectives and learn Microsoft's phrasing. Review why every alternative is wrong, especially when several tools could technically produce the same result.

Practice material should align with the public blueprint and must not reproduce live exam content. Avoid dumps and claims of “real questions.” They do not build the architecture and troubleshooting judgment DP-600 is designed to assess.

DP-600 compared with DP-700 and PL-300

Choose DP-600 when your work centers on enterprise Fabric analytics and semantic models. Choose DP-700 for data engineering, orchestration, ingestion, and platform optimization. PL-300 is the more Power BI-specific analyst path and is a better fit when Fabric lakehouses, warehouses, KQL, and enterprise lifecycle responsibilities are outside your role.

Is DP-600 worth it?

The certification is valuable when your organization uses Fabric for governed analytics or when a target role combines Power BI modeling with Fabric data assets. It signals broader scope than report creation alone. Its value is lower if you do not expect to work with Fabric, so pair the credential with a project that demonstrates the skills.

Exam-day notes

• You have 100 minutes for the assessment; administrative seat time is longer.
• Microsoft does not publish a fixed question count or promise specific question types.
• A scaled score of 700 or higher passes; it is not equivalent to 70% correct.
• Microsoft Learn is available during associate exams, but the timer continues.
• Use a personal Microsoft account when scheduling so the record remains yours.

Prep resources

← More Microsoft exams