Last updated: June 2026. First introduced as the successor path to the retired AZ-303 and AZ-304 exams, AZ-305 reflects Microsoft’s move away from split implementation-and-design testing toward a single architecture exam centred on design decisions.
The AZ-305 exam, Designing Microsoft Azure Infrastructure Solutions, is the required exam for the Microsoft Certified: Azure Solutions Architect Expert certification. It tests whether a candidate can design Azure and hybrid solutions that satisfy business requirements, technical constraints, security expectations, operational needs, and cost boundaries. The current Microsoft study guide for AZ-305, including skills measured and domain weightings, is published on Microsoft Learn and should be checked before scheduling the exam because the outline can change over time.
Editorial note: This article is periodically reviewed against Microsoft Learn exam pages, including the official AZ-305 study guide, certification page, and Microsoft exam policy guidance. Candidates should treat Microsoft’s pages as the authoritative source for live exam objectives, pricing, registration, and renewal details.
AZ-305 is not an administrator exam with a different title. It assumes the candidate already understands Azure administration, identity, networking, storage, compute, security, and monitoring well enough to make design decisions across them. The exam is concerned with why one design should be chosen over another, not just how to configure a single service in the portal.
This is the most important difference between AZ-305 and the retired AZ-303/AZ-304 route. The older path separated architecture and implementation more explicitly. AZ-305 places more weight on design-first judgement: governance, resilience, cost management, migration strategy, identity boundaries, and service selection under constraint. A candidate who memorises product names but cannot explain trade-offs will struggle with the scenario-based nature of the exam and with the work the certification represents.
The official skills outline groups the exam into four broad areas. Microsoft currently presents these domains as designing identity, governance, and monitoring solutions; designing data storage solutions; designing business continuity solutions; and designing infrastructure solutions. In real projects, those domains rarely appear separately. A storage choice affects network design, a recovery objective affects cost, and an identity decision may shape the whole management model.
The identity, governance, and monitoring domain is about control. In project terms, it asks whether an architect can design Microsoft Entra ID integration, role-based access control, management groups, Azure Policy, logging, and monitoring in a way that supports accountability without slowing every delivery team. A common weak point is treating governance as an afterthought. In mature Azure estates, policy guardrails, tagging requirements, diagnostic settings, and landing zone structure are part of the foundation rather than an audit clean-up activity.
The data storage domain tests selection and design across relational, non-relational, object, and analytical storage patterns. The practical skill is not simply knowing that Azure SQL Database, Cosmos DB, and Azure Storage exist. It is knowing when consistency, query model, latency, scale, data residency, backup requirements, or operational familiarity should drive the choice. For example, Cosmos DB may suit globally distributed low-latency applications, but its partitioning, consistency, and cost model need deliberate design; Azure SQL may be a stronger fit when relational integrity and existing operational skills matter more.
The business continuity domain focuses on resilience, backup, high availability, disaster recovery, and recovery planning. The exam can test whether a design meets recovery time and recovery point objectives, but the real judgement is in balancing resilience against cost and complexity. Active-active architectures can be attractive on paper, yet many organisations need simpler regional failover, clear runbooks, tested restore procedures, and monitoring that proves the design works under failure.
The infrastructure domain brings together compute, networking, migration, and application architecture. This includes choices such as App Service versus AKS versus Azure Functions, hub-spoke networking, hybrid connectivity, private access, load balancing, migration approaches, and platform operations. The domain rewards candidates who can reason about operational ownership. A container platform may be technically suitable, but if the team lacks Kubernetes operations maturity, App Service or Container Apps may offer a better balance of control and maintainability.
The AZ-305 exam maps to the Azure Solutions Architect Expert credential. It is usually pursued by practitioners who have already worked with Azure administration, development, infrastructure, or DevOps responsibilities. Microsoft does not position it as an entry-level certification, and the exam objectives assume fluency across several Azure service families.
Many candidates reach AZ-305 after Azure Fundamentals or the Azure Administrator Associate certification. AZ-104 is not a formal prerequisite for AZ-305, but administrator experience helps because architecture decisions depend on how Azure resources behave after deployment. Candidates who need to strengthen those fundamentals may find Azure Administrator (AZ-104) training useful before moving into design-heavy preparation.
Azure developers can also move successfully into AZ-305, especially if they have experience with application hosting, identity integration, messaging, data services, and deployment pipelines. The Azure Developer Associate certification covers a different role focus, but it can provide useful context for application architecture decisions in AZ-305. The gap for developers is often governance, networking, and enterprise operations rather than application design itself.
AZ-305 is delivered through Microsoft’s certification exam programme and is commonly scheduled through Pearson VUE from the Microsoft certification page. Registration, regional pricing, language availability, online proctoring options, and test-centre availability should be confirmed on Microsoft’s official pages at the time of booking, because these details can vary by location and change over time.
Microsoft certification exams use a scaled scoring model, with a passing score of 700. The number of questions can vary, and candidates may see question types such as multiple choice, drag-and-drop, case studies, build lists, and scenario-based items. The Microsoft exam sandbox is useful for becoming familiar with the interface and item formats without relying on memorised question dumps or material that may breach exam rules.
The seat time includes exam instructions, the exam itself, optional feedback, and any scheduled break shown in the exam experience. Candidates should read Microsoft’s current exam information carefully before test day, including how breaks work, because time management can affect performance on case studies. A practical approach is to answer straightforward items efficiently, reserve attention for case-study requirements, and avoid spending too long trying to infer details that the scenario does not provide.
Retake rules, cancellation windows, rescheduling rules, identification requirements, and online proctoring conditions are governed by Microsoft and its exam delivery partners. The same applies to renewal: Microsoft role-based and specialty certifications can require periodic renewal, usually through Microsoft Learn, and the certification page should be checked for current renewal requirements. Candidates should plan for renewal as part of ongoing professional practice rather than treating the exam as a one-time event.
The strongest AZ-305 preparation connects exam domains to real design trade-offs. For compute, the decision is rarely “which service is newest?” A workload with predictable web traffic, modest operational complexity, and a standard deployment pipeline may fit App Service. A highly event-driven process may fit Azure Functions if execution duration, state, networking, and observability requirements are well understood. AKS can be appropriate for microservices and portability needs, but it introduces cluster operations, upgrade planning, security boundaries, ingress design, and platform team responsibilities.
Data decisions follow the same pattern. Azure SQL Database often suits relational systems that need transactions, familiar SQL tooling, and managed database operations. Cosmos DB is compelling for globally distributed applications with suitable access patterns, but the partition key and consistency model can decide whether the design succeeds. Azure Storage is effective for objects, files, queues, and low-cost retention patterns, but it is not a relational database replacement. Good architecture work starts with workload behaviour, not with a service catalogue.
Networking is where many otherwise strong candidates lose design clarity. Private Endpoints improve private access to platform services, but they also introduce private DNS design requirements, especially across hub-spoke and hybrid networks. An architecture that works in a single virtual network may fail when name resolution crosses on-premises DNS, Azure Private DNS zones, and multiple subscriptions. AZ-305 candidates should be able to reason through how traffic flows, how DNS resolves, and where inspection or forced tunnelling is required.
Identity design also creates architectural boundaries. Microsoft Entra ID tenant structure, external collaboration, privileged access, conditional access, managed identities, and role assignments determine who can administer, deploy, and operate resources. In enterprise environments, the wrong management group hierarchy or overly broad role assignment can become a long-term security and governance problem. The exam may ask this through scenarios, but the underlying lesson is practical: access control should be designed before scale makes it difficult to correct.
Cost is another design constraint, not a finance appendix. Architects are expected to design for budgets, tagging, cost allocation, reserved capacity where appropriate, and lifecycle policies for data. A technically elegant design that ignores monitoring costs, cross-region replication, egress charges, over-provisioned compute, or duplicate logging is incomplete. This is also why AZ-305 preparation should include cost estimation and governance exercises rather than only service overviews.
Preparation should start with the official Microsoft skills measured page, then move quickly into design practice. Reading documentation is necessary, but it is not enough. Candidates need to practise turning requirements into architecture decisions, documenting assumptions, and defending trade-offs in the same way they would during a design review.
A useful study rhythm is to take one domain at a time and build a small end-to-end design around it. For governance, design an Azure landing-zone-style hierarchy with management groups, subscriptions, Azure Policy assignments, role assignments, diagnostic settings, and tagging rules. For networking, build a hub-spoke topology, add Private Endpoints, configure private DNS, and test connectivity from both Azure and a simulated hybrid endpoint. For resilience, deploy a workload with backup, availability, monitoring, and a failover runbook, then test what actually happens when a dependency is unavailable.
Infrastructure as code should be part of the preparation. Bicep or Terraform forces design choices to become repeatable, reviewable, and testable. It also exposes details that are easy to miss in diagrams, such as naming standards, dependency ordering, module boundaries, identity permissions, policy exemptions, and environment promotion. A candidate does not need to become a full-time platform engineer to pass AZ-305, but an architect who cannot read or reason about infrastructure code is at a disadvantage in modern Azure teams.
One common preparation mistake is studying isolated services in long lists. A stronger approach is to create complete designs that include IAM, networking, compute, data, monitoring, security, disaster recovery, and cost controls. Designs should be checked against the Azure Well-Architected Framework and Azure Landing Zone guidance so that preparation reflects current architectural patterns rather than one-off lab builds.
Hiring processes often reinforce the same point. Candidates for Azure architect roles may be asked to whiteboard a workload design, estimate cost drivers, explain trade-offs, and produce a short architecture decision record. AZ-305 preparation becomes more useful when it mirrors those expectations. Instead of only asking “what does this service do?”, candidates should ask “why is this service appropriate here, what are the risks, and what would make the decision change?”
Structured training can help when it adds guided labs, architecture discussion, and accountability rather than replacing independent design practice. Readynez covers AZ-305 through a Microsoft Azure Solutions Architect path, but candidates should still anchor their preparation in Microsoft Learn, hands-on lab work, and their own architecture reviews. A practical option for guided preparation is the Azure Solutions Architect (AZ-305) course with hands-on labs.
Practice tests are useful when they reveal weak reasoning. They are much less useful when they become a memorisation exercise. AZ-305 scenarios often include more information than is needed, and part of the skill is identifying the constraint that matters: compliance, latency, operational simplicity, recovery objective, existing licensing, or team capability.
After each practice question, the important work is reviewing why the correct answer fits and why the alternatives do not. If a candidate cannot explain the trade-off in plain language, the answer has not been learned deeply enough. Exam dumps and copied live-question material should be avoided because they create a false sense of readiness and can violate certification rules.
The value of AZ-305 preparation is that it pushes practitioners to think across service boundaries. Azure architecture is rarely a single-product problem. A workload design may need private connectivity, identity federation, database replication, monitoring, cost controls, incident response, and deployment automation before it is ready for production.
The most effective next step is to use the exam objectives as a design syllabus. Build small reference architectures, write short decision records, estimate costs, and review designs against security, reliability, operational, performance, and cost principles. Candidates who want a structured route can use Readynez as one preparation option, while keeping Microsoft’s official exam pages and hands-on Azure practice at the centre of the plan.
AZ-305 measures the ability to design Azure infrastructure solutions across identity, governance, monitoring, data storage, business continuity, and infrastructure. Microsoft publishes the live skills outline in the official AZ-305 study guide, and candidates should review that page before booking because exam objectives can be updated.
AZ-104 is not listed as a formal prerequisite for AZ-305, but Azure administration experience is highly useful. AZ-305 assumes candidates can reason about subscriptions, networking, storage, compute, identity, monitoring, and governance from an operational perspective, which is why many candidates complete or study AZ-104 first.
The exam can include several Microsoft exam item types, including multiple-choice questions, case studies, drag-and-drop items, build lists, and scenario-based questions. Candidates should use the Microsoft exam sandbox to understand the interface and should confirm current seat time, registration details, and exam policies on Microsoft’s official certification pages.
Good lab preparation should cover complete designs rather than isolated services. Useful exercises include building a landing-zone-style management group and policy structure, deploying a hub-spoke network with Private Endpoints and private DNS, designing backup and failover for a workload, comparing compute options, and modelling cost implications for different architecture choices.
Get Unlimited access to ALL the LIVE Instructor-led Microsoft courses you want - all for the price of less than one course.
You're viewing our global site from United States
Would you like to view the site in
English
with prices in
Dollar?