PCA study guide: Professional Cloud Architect domains and a 6-week plan

The Google Cloud Professional Cloud Architect (PCA) exam tests whether you can design, build and run secure, scalable, cost-effective solutions on Google Cloud that serve real business goals. It rewards judgement over recall: most questions describe a company, its constraints and a goal, then ask which design fits best. You answer 50 to 60 multiple-choice and multiple-select questions in two hours, and some of them refer to published case studies you can read before exam day.

PCA at a glance

Details as of September 2026 — confirm on the official exam page before booking.

Each exam includes two case studies, and Google states that case study questions make up 20–30% of the exam. The current exam guide lists four case studies that may be used: Altostrat Media, Cymbal Retail, EHR Healthcare and KnightMotives Automotive. Google notes the exam was updated for recent branding changes.

Who this exam is for

Google recommends three or more years of industry experience, including at least one year designing and managing solutions on Google Cloud. There is no formal prerequisite. The exam guide describes an architect who is proficient in enterprise cloud strategy, solution design, workload migration, deployment and orchestration, and who is comfortable with open-source technologies and multitiered applications across legacy, hybrid and multicloud environments.

The guide is explicit that familiarity with the Google Cloud Well-Architected Framework is a key requirement. Its pillars (operational excellence, security, reliability, performance optimisation, cost optimisation and sustainability) run through every domain, so treat the framework as the lens for every answer.

If you are newer to Google Cloud, the [ACE study guide](/blog/ace-study-guide) covers the Associate Cloud Engineer exam, which builds the hands-on base PCA assumes.

What the exam covers

The guide has six sections. Section 1 carries a quarter of the exam on its own, and the first three sections together make up 60%.

Section 1: Designing and planning a cloud solution architecture (~25%)

This is the heart of the exam. It asks you to turn business requirements (continuity, cost, integration, success measures such as KPIs and ROI, compliance, observability) and technical requirements (availability, failover, scalability, latency, backup and recovery) into a design. You need to choose networking, storage, compute and data processing options, including when GKE, Cloud Run or Cloud Run functions fit, and when spot VMs or custom machine types make sense. The section also covers AI options such as Gemini models, Agent Builder, Model Garden and AI Hypercomputer, Gemini Cloud Assist, migration planning with Google Cloud Migration Center, and workload disposition (build, buy, modify or deprecate). The judgement tested is the trade-off: the design that meets every stated requirement with the least cost and complexity.

Section 2: Managing and provisioning a cloud solution infrastructure (~17.5%)

Here the focus shifts to configuration. Expect hybrid and multicloud networking, VPC design, load balancing and firewall protection; storage allocation, lifecycle, retention and data protection; and compute provisioning across Compute Engine, GKE, serverless and Google Cloud VMware Engine, including spot versus standard capacity. This section also covers Gemini Enterprise Agent Platform: automating the ML lifecycle with Agent Platform Pipelines, preparing data integration, using GPUs and TPUs through AI Hypercomputer, choosing between the Google AI APIs, integrating Gemini Enterprise features such as AI agents and NotebookLM, and bringing Model Garden models into a solution.

Section 3: Designing for security and compliance (~17.5%)

Security questions cover IAM, the resource hierarchy, separation of duties, key and secret management (including customer-managed keys in Cloud KMS), and controls such as VPC Service Controls, organisation policy, context-aware access and hierarchical firewall policies. Secure remote access brings in Identity-Aware Proxy, service account impersonation, Chrome Enterprise Premium and Workload Identity Federation. The guide adds software supply chain security and securing AI with Model Armor and Sensitive Data Protection. Compliance questions tie these to regulation (health records, children's privacy, data sovereignty), commercial data such as card details and PII, industry certifications such as SOC 2, and audit logging.

Section 4: Analyzing and optimizing technical and business processes (~15%)

This section tests the architect as adviser. Technical processes include the software development lifecycle, CI/CD, root cause analysis, testing, service catalogues and disaster recovery. Business processes include stakeholder and change management, team skills readiness, decision-making, customer success, CapEx versus OpEx and business continuity. The right answer respects the organisation's skills, budget and change tolerance.

Section 5: Managing implementation (~12.5%)

Expect questions on advising development and operations teams: deployment approaches, API management with Apigee, load, unit and integration testing, migration tooling and Gemini Cloud Assist. The second half covers working with Google Cloud programmatically: Cloud Shell and Cloud Code, the gcloud, gsutil and bq tools, emulators for Bigtable, Spanner, Pub/Sub and Firestore, Terraform and infrastructure as code, and good practice for Google APIs and client libraries.

Section 6: Ensuring solution and operations excellence (~12.5%)

The final section rests on the operational excellence pillar, and covers Google Cloud Observability (monitoring, logging, profiling and alerting strategies), deployment and release management, supporting live solutions, quality control, and reliability practices such as chaos engineering, penetration testing and load testing.

A 6-week study plan

The plan follows the weights, starts with the case studies and ends with a week of mocks. Our [study plan template](/blog/cloud-certification-study-plan-template) shows how to adapt it to your own calendar.

Week 1: Framework and case studies. Read the Well-Architected Framework pillars and summarise each in a paragraph. Then read all four case studies and, for each, list the business requirements, technical requirements and constraints in a table. Revisit these tables every week.

Week 2: Section 1, design. Work through compute choice (Compute Engine, GKE, Cloud Run, Cloud Run functions), storage and database choice, networking patterns and data processing. For each, write the requirement that points to it. Take a Section 1 domain paper at the end of the week.

Week 3: Section 1 continued and Section 2. Cover migration planning, high availability and disaster recovery designs, then hybrid networking, storage lifecycle and compute provisioning. Spend two sessions on Agent Platform and the AI services, learning each by the problem it solves. Take a Section 2 paper.

Week 4: Section 3, security and compliance. Map every control in the guide to the threat it addresses: IAM and the hierarchy for access, VPC Service Controls for data exfiltration, Cloud KMS for key ownership, IAP and Workload Identity Federation for access without keys. Take a Section 3 paper and review every explanation.

Week 5: Sections 4, 5 and 6. These carry 40% between them, but the material is lighter on product detail. Focus on CI/CD, testing, release strategies, observability and reliability practices, and practise the business-process questions by asking what the stakeholder actually needs. Take a paper for each section.

Week 6: Mocks and review. Sit a full-length mock under exam conditions, two hours without breaks. Review every question and re-read the case studies against your mistakes. Fix the weakest areas, then sit a second mock two or three days later.

Common traps

• Answering the technology, not the requirement. A design that is impressive but ignores a stated constraint (budget, a small team, a data residency rule) is wrong. Underline the constraint before reading the options.
• Over-engineering. When two options both work, the managed, simpler one usually fits the Well-Architected Framework better.
• Reading case studies only in the exam. Case study questions are 20–30% of the exam. Knowing the four companies' goals in advance saves time and stops you missing a requirement buried in the text.
• Mixing up similar security controls. IAM decides who can act, organisation policy restricts what can be configured, VPC Service Controls limit where data can move, and firewall policies control traffic. Distractors are often real controls doing the wrong job.
• Using old product names. The guide uses Gemini Enterprise Agent Platform, Cloud Run functions, Chrome Enterprise Premium and Google Cloud Observability. If your notes use older names, map them before you study further.
• Ignoring the business sections. Sections 4 and 5 are easy to skip for technical candidates, but together they are 27.5% of the exam.

How to practise

CertifyCloudx has original PCA practice questions written against the current exam guide, each explained option by option. Read [how our questions are written](/blog/how-certifycloudx-practice-questions-are-written) for the method, and [how to use practice exams effectively](/blog/how-to-use-practice-exams-effectively) to get the most from each attempt.

• Full-length timed mock exams with the real exam's two-hour limit
• Domain papers of up to 25 questions each (60 minutes per 25)
• Original case studies in the style of the exam, with linked questions
• Progress tracked by domain and topic

The free plan includes practice sets for every certification, up to 10 questions a day, with no card required. Start with the [PCA practice questions](/certifications/gcp-professional-cloud-architect).

Frequently asked questions

How difficult is the Professional Cloud Architect exam?

It is a professional-level exam that assumes three or more years of industry experience, including a year designing solutions on Google Cloud. The difficulty comes from scenario questions where several answers are technically possible and only one fits every stated requirement.

How long should I prepare for PCA?

With solid Google Cloud experience, six weeks of steady study is a realistic plan. Without associate-level experience, allow longer.

Which case studies are used on the PCA exam?

The current exam guide lists four: Altostrat Media, Cymbal Retail, EHR Healthcare and KnightMotives Automotive. Each exam includes two of them, and case study questions make up 20–30% of the exam. Read the versions linked from the official exam guide, as Google updates them.

How long is the certification valid, and how do I renew it?

The certification is valid for two years. Google offers a renewal exam for active certification holders within the renewal eligibility period: 25 questions in one hour for $100 plus tax, built around one generative AI case study.

Are CertifyCloudx questions real PCA exam questions?

No. Every CertifyCloudx question and case study is original and written against the public exam guide. Relying on leaked exam content can put your certification at risk.

CertifyCloudx is independent and not affiliated with Google. Google Cloud Certified – Professional Cloud Architect is a trademark of its owner. All CertifyCloudx practice questions are original.