Microsoft Azure Data Fundamentals
175 practice questions
Last reviewed: April 2026
Personal notes and resource links for your study journey
Filter by Certification
Microsoft Azure Data Fundamentals (DP-900) is the entry-level credential for data workloads on Azure. It targets candidates beginning a data-platform career — junior data engineers, analysts, BI developers, and stakeholders who need fluency in Azure data services. The exam validates conceptual understanding of relational and non-relational data, batch and streaming analytics workloads, and the Azure portfolio that supports each (Azure SQL family, Cosmos DB, Storage, Synapse, Data Factory, Microsoft Fabric, Databricks). Expect 40–60 multiple-choice, multiple-response, and drag-and-drop questions in 45 minutes — conceptual, not hands-on, with no case studies.
Structured / semi-structured / unstructured data, OLTP vs. OLAP, batch vs. streaming, common data roles (engineer, analyst, scientist) and the modern data warehouse / lakehouse / data mesh patterns. About 28% of questions.
Azure SQL Database vs. SQL Managed Instance vs. SQL Server on VMs, Azure Database for MySQL / PostgreSQL / MariaDB, basic SQL, normalization, and managed-service tradeoffs. About 22%.
Azure Cosmos DB APIs (NoSQL, MongoDB, Cassandra, Gremlin, Table), Azure Storage (Blob, Files, Tables, Queues), document / key-value / graph / column-family data models. About 18%.
Largest domain at 32%. Azure Synapse Analytics, Microsoft Fabric, Azure Databricks, Azure Data Factory, Power BI, plus core warehouse / lakehouse architecture concepts. Heaviest growth area in recent refreshes.
Services you'll encounter on the exam and why each one matters.
Fully managed PaaS relational engine built on the latest stable SQL Server, with built-in HA, automatic patching, and elastic scale.
Why it's on the exam: The headline PaaS option in Domain 2 (Relational Data on Azure) — expect questions distinguishing it from Managed Instance and SQL on VMs by management overhead.
PaaS deployment of a near-100% feature-compatible SQL Server instance, with SQL Agent, cross-database queries, and CLR support.
Why it's on the exam: Domain 2 tests when to choose Managed Instance over Azure SQL Database for lift-and-shift workloads needing instance-level features.
IaaS deployment of SQL Server on Windows or Linux VMs, giving full OS-level and SQL-engine control with optional automation extensions.
Why it's on the exam: Domain 2 frames it as the maximum-control, maximum-responsibility relational option — contrasted with managed PaaS tiers.
Managed open-source relational database services with flexible-server deployment, automated backups, and built-in HA.
Why it's on the exam: Domain 2 names the open-source PaaS family as the answer when the workload is already on Postgres or MySQL — distinguish from SQL Server tiers.
Globally distributed multi-model NoSQL database supporting key-value, document, column-family, graph, and vector APIs with tunable consistency.
Why it's on the exam: The flagship answer across Domain 3 (Non-Relational Data) — expect questions on API choice, partitioning, RU/s, and consistency-level tradeoffs.
Umbrella account hosting Blob, File, Queue, and Table services with redundancy options (LRS/ZRS/GRS) and access tiers (hot/cool/cold/archive).
Why it's on the exam: Domain 3 tests Blob/Files/Tables as the canonical non-relational stores; Domain 1 tests storage tiers and redundancy as core concepts.
Blob storage with a hierarchical namespace and POSIX ACLs optimized for big-data analytics workloads in Synapse, Databricks, and Fabric.
Why it's on the exam: Domain 4 (Analytics Workloads) treats ADLS Gen2 as the default lake-storage substrate that analytics services consume.
Unified analytics platform combining OneLake (a single SaaS data lake), Data Factory, Synapse Engineering/Warehousing/Real-Time, and Power BI under one capacity.
Why it's on the exam: Domain 4 leads with Fabric as the current end-to-end analytics story — expect questions on OneLake, workloads, and the shortcut/lakehouse model.
Enterprise analytics service unifying dedicated/serverless SQL pools, Apache Spark pools, and pipelines for warehouse + big-data workloads.
Why it's on the exam: Domain 4 distinguishes Synapse from Fabric and Databricks — questions test when a dedicated SQL pool fits versus serverless or Spark.
Managed Apache Spark and Delta Lake platform for big-data engineering, ML, and lakehouse analytics co-developed with Databricks.
Why it's on the exam: Domain 4 names Databricks as the Spark-centric lakehouse choice; expect contrast questions against Synapse Spark pools and Fabric.
Cloud-native ETL/ELT orchestrator with 90+ connectors, mapping data flows on Spark, and integration runtimes for hybrid scenarios.
Why it's on the exam: Domain 4 tests Data Factory as the canonical pipeline orchestrator for moving and transforming data across stores.
Real-time, SQL-like stream processing engine with built-in temporal windows, joining Event Hubs/IoT Hub inputs to Power BI, Synapse, or storage sinks.
Why it's on the exam: Domain 4 questions on real-time analytics name Stream Analytics as the low-code answer for windowed aggregations on event streams.
Highly scalable event ingestion service handling millions of events per second with a Kafka-compatible endpoint and capture-to-storage.
Why it's on the exam: Domain 4 frames Event Hubs as the front door for streaming ingestion feeding Stream Analytics, Fabric Real-Time, or Databricks.
Self-service BI platform for interactive reports, semantic models, and dashboards over Fabric, Synapse, SQL, and dozens of other sources.
Why it's on the exam: Domain 4 tests Power BI as the visualization layer in modern analytics architectures — expect questions on workspace, dataset, and report concepts.
Fabric workload built on the Kusto/KQL engine for ingesting, exploring, and acting on high-velocity event data via Eventstreams and Eventhouses.
Why it's on the exam: Domain 4 increasingly names Real-Time Intelligence as the Fabric-native answer for streaming analytics, displacing standalone HDInsight scenarios.
Managed in-memory key-value store based on Redis, supporting caching, session state, and pub/sub for low-latency reads.
Why it's on the exam: Domain 3 cites Azure Cache for Redis as the canonical in-memory NoSQL example — contrast with persistent NoSQL stores like Cosmos DB.
Cloud identity and access management service providing authentication, conditional access, and managed identities for Azure data services.
Why it's on the exam: Domain 1 + Domain 2 questions on securing data access name Entra ID as the identity plane behind Azure SQL AAD auth and storage RBAC.
Unified data governance service for cataloguing, classifying, and tracking lineage across Azure, on-prem, and multi-cloud data estates.
Why it's on the exam: Domain 1 (Core Data Concepts) tests Purview as the catalog/lineage answer when a scenario asks how to inventory and govern enterprise data.
Managed service for storing secrets, keys, and certificates with HSM-backed options, used to encrypt data and broker DB credentials.
Why it's on the exam: Data-protection scenarios across Domains 2–4 cite Key Vault for TDE customer-managed keys, connection-string secrets, and certificate management.
Platform telemetry service collecting metrics, logs, and traces from data services, with Log Analytics (KQL) and alert rules.
Why it's on the exam: Domain 4 expects Azure Monitor for surfacing query performance, pipeline failures, and storage throttling across the analytics estate.
$65k–$100k–$140k USD annual
DP-900 by itself does not move salaries materially — it is a literacy signal. Engineers entering the higher end pair it with DP-203 / DP-300 / DP-600 / DP-700 plus 2–4 years of hands-on data work. Non-coastal US markets trend toward the lower end.
Source: levels.fyi 2025 data analyst / engineer roles, U.S. BLS OEWS May 2024 (15-2051 data scientists, 13-2031 budget / data analysts), Glassdoor 2025. Figures are approximate; actual compensation depends on role, region, and experience.
DP-900 has had steady demand as enterprises modernize data estates onto Azure Synapse, Databricks, and increasingly Microsoft Fabric. Recruiters treat it as a baseline literacy signal for analyst and junior engineer roles, and as evidence of intent for engineers transitioning into data platform work. It pairs naturally with AZ-900 to round out an Azure-platform overview, and acts as the recommended on-ramp to the associate-level data exams (DP-300 for DBAs, DP-600 / DP-700 for Fabric, DP-100 for data science). Microsoft regularly distributes free DP-900 vouchers through Microsoft Learn data-skills challenges.
There are no formal prerequisites. DP-900 is positioned as the first data-track exam most candidates take. Microsoft's free Microsoft Learn path covers all four domains in roughly 10–15 hours of self-paced content, and the included sandboxes let you experiment with Azure SQL, Cosmos DB, and Synapse without a paid subscription.
If you already hold AZ-900 or AI-900, plan to compress study time meaningfully — only the data-specific domains are new. If you are coming from AWS Cloud Practitioner or AWS Data Engineer Associate, focus on mapping AWS service names (RDS, DynamoDB, Glue, Redshift) onto Azure equivalents (Azure SQL, Cosmos DB, Data Factory, Synapse / Fabric).
DP-900 sits in the Fundamentals tier — comparable in difficulty to AZ-900 and AI-900. Plan on 15–25 hours of study over 2–3 weeks with no prior data background; experienced data professionals often pass with 5–10 hours of focused review. The exam runs 45 minutes with 40–60 questions in mixed formats: multiple choice, multiple response, and drag-and-drop matching exercises. No case studies at the Fundamentals tier.
The most common stumbling block is service overlap — Azure has multiple ways to land a data-warehouse, lakehouse, or streaming workload (Synapse vs. Fabric vs. Databricks vs. Stream Analytics) and the exam expects you to know which is the canonical choice for each scenario. Microsoft Fabric coverage has grown significantly in the 2024 outline refresh, so older study material may underweight it.
Refreshed to add Microsoft Fabric coverage in the analytics domain and update Azure Synapse / Databricks framing. Microsoft refreshes DP-900 approximately yearly without changing the exam code.
Pre-Microsoft-Fabric outline. Heavier on Synapse and Data Factory; no Fabric or Lakehouse-by-name coverage.
DP-900 (Microsoft Azure Data Fundamentals) is a considered an entry-level exam testing breadth of conceptual understanding rather than hands-on depth Foundational-level exam. Most candidates need 30–80 hours of study spread over 3–6 weeks for foundational-level exams. Most candidates who score consistently above the passing threshold on practice exams pass on their first attempt.
Most candidates need 30–80 hours of study spread over 3–6 weeks for foundational-level exams. Time-to-pass varies widely by prior experience. Engineers with hands-on production experience in the underlying technology typically need less; candidates new to the platform should plan toward the upper end of that range.
DP-900 is a recognized credential in the Azure ecosystem and signals validated knowledge to employers, recruiters, and clients. Whether it is worth the time and fee for you depends on your role and goals — it tends to pay off most for cloud engineers, architects, and consultants who work with Azure day-to-day or want to move into roles that do.
The passing score for DP-900 is 700 / 1000. The exam contains 40 questions and lasts 45 min.
The DP-900 exam fee is $99 USD. Fees are set by Azure and may vary by region; always confirm the current price on the official Azure certification page before booking.
Microsoft fundamentals certifications never expire (AZ-900, AI-900, DP-900, SC-900).
Yes. You can take the exam online (proctored via the provider's secure browser, available 24/7 in most regions) or at an in-person Pearson VUE test center during business hours. Both formats use the same questions, time limit, and passing score.
CertLabPro provides 15 study modes across the practice question bank for DP-900. The exam-simulation mode mirrors the real exam: 40 questions in 45 min, with the same passing threshold of 700 / 1000. Browse mode lets you read every Q&A statically.