Google Cloud Professional Cloud Developer
225 practice questions
Last reviewed: April 2026
Personal notes and resource links for your study journey
Filter by Certification
The Google Cloud Professional Cloud Developer (PCD) validates the ability to build scalable, secure, cloud-native applications on Google Cloud. The exam emphasizes real engineering: choosing between Cloud Run, GKE, App Engine, and Cloud Functions for a given workload; designing event-driven systems with Pub/Sub and Eventarc; implementing observability with Cloud Trace and Cloud Profiler; and integrating with Cloud SQL, Spanner, Firestore, and Memorystore. Question style is scenario-heavy — short stories about a team's requirements with four plausible-looking options where one is the most idiomatic Google Cloud answer. PCD is the GCP analog of AWS Developer Associate stepped up a tier, or Azure AZ-204 with deeper architecture content. It targets backend developers, full-stack engineers, and platform engineers who actually ship code on GCP.
Largest domain at 25%. 12-factor patterns, microservices vs. monolith tradeoffs, choosing the right compute (Cloud Run vs. GKE vs. Cloud Functions vs. App Engine), event-driven design with Pub/Sub. Heavy on tradeoff scenarios.
Local development with Cloud Code, container builds with Cloud Build and Buildpacks, unit / integration / load testing patterns, dependency management. 20%.
Blue-green and canary deploys on Cloud Run and GKE, traffic splitting, Artifact Registry, deployment-pipeline patterns with Cloud Build, service identity and Workload Identity. 20%.
Pub/Sub patterns (push vs. pull, exactly-once delivery), Eventarc, Cloud Tasks, Cloud Scheduler, integrations with Cloud SQL / Spanner / Firestore / Memorystore. 20% — heavy on idiomatic patterns.
Cloud Logging, Cloud Monitoring, Cloud Trace, Cloud Profiler, error reporting, autoscaling tuning, cost-aware engineering. 15%.
Services you'll encounter on the exam and why each one matters.
Fully managed serverless container runtime that scales to zero, with revisions, traffic splitting, concurrency tuning, and request-based billing.
Why it's on the exam: Domain 1 (Designing Cloud-Native Applications) treats Cloud Run as the default compute primitive for stateless containerised services — expect questions on concurrency, cold starts, and revision rollback.
Event-driven serverless compute (2nd gen built on Cloud Run) triggered by HTTP, Pub/Sub, Cloud Storage, Eventarc, and Firestore events.
Why it's on the exam: Domain 1 + Domain 4 (Integrating Google Cloud Services) test Functions for lightweight glue between services and event-source mapping patterns.
Fully managed PaaS for web apps with Standard (per-request scaling) and Flexible (container-based) environments, traffic splitting, and versioned deployments.
Why it's on the exam: Domain 3 (Deploying Applications) tests App Engine traffic splits, versioned deploys, and Standard-vs-Flexible selection — a recurring distractor against Cloud Run.
Managed Kubernetes with Autopilot and Standard modes, Workload Identity, multi-cluster ingress, and integrated Anthos service mesh.
Why it's on the exam: Domain 1 + Domain 3 test GKE Autopilot vs. Standard tradeoffs, Workload Identity for pod-to-Google-API auth, and rolling-vs-blue/green deployment strategies.
Managed CI service driven by `cloudbuild.yaml` that builds container images, runs tests, signs artifacts, and pushes to Artifact Registry.
Why it's on the exam: Domain 2 (Building and Testing Applications) names Cloud Build as the canonical CI service — expect questions on build steps, substitutions, and triggers.
Managed private Git repositories integrated with Cloud Build triggers, Cloud Logging, and IAM-based repo access.
Why it's on the exam: Domain 2 + Domain 3 cover CSR as the Google-native source for Cloud Build triggers, often contrasted with mirrored GitHub/GitLab repos.
Managed continuous-delivery service for GKE, Cloud Run, and Anthos with declarative delivery pipelines, promotion gates, and automated rollback.
Why it's on the exam: Domain 3 (Deploying Applications) tests Cloud Deploy pipeline progression (dev → staging → prod), approval gates, and rollback semantics.
Unified storage for container images, Maven, npm, Python, Go, and OS packages with IAM-controlled access, vulnerability scanning, and remote/virtual repos.
Why it's on the exam: Domain 2 + Domain 3 expect Artifact Registry (successor to Container Registry) as the image store consumed by Cloud Build → Cloud Run / GKE deploys.
Serverless document database (Firestore in Native or Datastore mode) with real-time listeners, offline sync, and security rules; Firebase Realtime DB for legacy JSON-tree apps.
Why it's on the exam: Domain 1 tests Firestore for low-latency document workloads with real-time push to mobile/web clients — a recurring contrast against Cloud Spanner and Bigtable.
Globally distributed, strongly consistent relational database with horizontal scaling, regional/multi-region instances, and SQL + GoogleSQL dialect support.
Why it's on the exam: Domain 1 names Spanner whenever a question demands relational semantics at planet scale or 5-nines availability — distinguishes from Cloud SQL and Firestore.
Global asynchronous messaging with at-least-once delivery, push and pull subscriptions, ordering keys, dead-letter topics, and Pub/Sub Lite for high-throughput regional workloads.
Why it's on the exam: Domain 4 (Integrating Google Cloud Services) tests Pub/Sub as the canonical decoupling primitive for event-driven backends — choosing pull vs. push and DLQ semantics are common questions.
Managed asynchronous task queues with per-task scheduling, rate/concurrency control, retries with exponential back-off, and HTTP / App Engine targets.
Why it's on the exam: Domain 4 distinguishes Cloud Tasks (explicit, deduplicated per-task control) from Pub/Sub (fan-out streams) — a recurring DVA-style distractor pair.
Serverless workflow orchestrator using a YAML/JSON DSL to chain HTTP, Cloud Functions, Cloud Run, and Google Cloud APIs with retries and parallel steps.
Why it's on the exam: Domain 4 contrasts Workflows (durable orchestration) with raw Pub/Sub fan-out — questions test step retries, parallel branches, and connector use.
Apigee provides full lifecycle API management (proxies, monetization, analytics, OAuth); API Gateway is a lightweight managed gateway for serverless backends.
Why it's on the exam: Domain 4 tests Apigee for enterprise API lifecycle (developer portal, quota, OAuth proxies) and API Gateway for serverless API exposure — choosing between them is a recurring scenario.
OpenAPI/gRPC API management layer for App Engine, GKE, and Compute Engine backends with ESP/ESPv2 proxy, authentication, monitoring, and quota enforcement.
Why it's on the exam: Domain 4 keeps Cloud Endpoints in scope for self-hosted backends on GKE/Compute Engine where Apigee is overkill but API Gateway doesn't fit.
Eventing infrastructure for delivering 130+ Google Cloud event sources to Cloud Run, Cloud Functions, and GKE via CloudEvents-formatted Pub/Sub or audit-log triggers.
Why it's on the exam: Domain 4 names Eventarc whenever an event must flow from one Google Cloud service (e.g. Cloud Storage, Audit Logs) to a serverless target with standard CloudEvents formatting.
Project- and organization-scoped identity with predefined and custom roles, service accounts, Workload Identity Federation, and resource-based policies.
Why it's on the exam: Domain 5 (Managing Application Performance / Security) tests least-privilege role design, service-account impersonation, and Workload Identity for pod-to-Google-API auth without static keys.
Cloud KMS manages customer-managed encryption keys (CMEK / external HSM); Secret Manager stores API keys, DB credentials, and tokens with versioning and rotation.
Why it's on the exam: Domain 1 + Domain 3 expect Secret Manager for runtime credential injection and Cloud KMS as the backing key store — distinguishing them from environment variables is a recurring question.
Cloud Trace for distributed latency analysis, Cloud Profiler for production CPU/memory profiling, and Cloud Debugger (deprecated; Snapshot Debugger replacement) for live-state inspection.
Why it's on the exam: Domain 5 (Managing Application Performance) is largely Cloud Operations Suite — expect questions on tracing chains across Cloud Run/Functions and profile-driven hotspot diagnosis.
Cloud Logging aggregates structured logs from every Google Cloud service with sinks, log-based metrics, and Logs Explorer; Error Reporting groups and alerts on application exceptions.
Why it's on the exam: Domain 5 tests structured logging from Cloud Run/Functions, log-based alerting via Cloud Monitoring, and Error Reporting groupings — the canonical "where is the bug" surface.
$130k–$180k–$270k USD annual
Range reflects US-based senior backend / cloud-native engineers where GCP is the primary platform. FAANG L5 software engineer TC clears $300k. The cert is a strong signal but does not by itself unlock these salaries — it complements 5–10+ years of demonstrated software engineering experience.
Source: levels.fyi 2025–2026 (Google L4–L5 software engineers, FAANG and GCP-shop unicorn senior backends), U.S. BLS OEWS May 2024 (15-1252 software developers). Figures are approximate; actual compensation depends on role, region, and experience.
PCD is requested less often than the architect-track certs but is a strong differentiator on senior backend and platform-engineering job postings at GCP-heavy companies. Demand concentrates at Spotify, Snap, PayPal, Wayfair, several major retailers, gaming studios, and Google Cloud partners. The cert is also valued at Google itself — customer-engineering and developer-advocate ladders frequently list it as preferred. PCD pairs naturally with the Kubernetes CKAD certification and with Terraform Associate to form a strong cloud-native developer profile. It is less commonly used as a hiring filter than ACE or PCA, but holders consistently report better recruiter response on senior-engineer roles.
There are no formal prerequisites. Google recommends three or more years of industry experience including one or more years designing and developing applications on Google Cloud. In practice, PCD is not a sensible first GCP cert for non-developers — successful candidates can read and write Go, Java, Python, or Node.js comfortably and have shipped non-trivial applications.
The Associate Cloud Engineer (ACE) is the most common stepping stone but is not strictly required if you already write production code on AWS or Azure. Comfort with containers, basic Kubernetes (Deployments, Services, ConfigMaps), CI/CD concepts, and at least one of the major SQL or document databases is effectively required. The official Cloud Developer Learning Path on Google Cloud Skills Boost (around 50–70 hours of labs and reading) is a good baseline, but most successful candidates supplement with personal Cloud Run / GKE side projects.
PCD is rated professional and is genuinely scenario-heavy. Plan on 80–130 hours of study over 8–12 weeks if PCD is your first GCP professional cert, or 40–70 hours over 4–6 weeks if you already hold ACE plus solid backend engineering experience. The exam is 50–60 multiple-choice / multiple-select questions in 120 minutes, delivered through Pearson VUE (Google migrated from Kryterion / Webassessor in early 2026).
The most common stumbling block is choosing between Cloud Run, GKE, App Engine, and Cloud Functions for a given scenario — Google's "preferred" answer often hinges on subtle scaling, latency, or operational-overhead criteria that are not obvious from documentation alone. The second stumbling block is Pub/Sub delivery semantics (at-least-once vs. exactly-once, push vs. pull, dead-letter topics). Google does not publish numeric scores — only pass/fail. The credential is valid for two years and recertification requires re-passing the current exam.
Current exam guide refreshed in late 2023 to add Cloud Run jobs, expanded Eventarc coverage, and updated GKE Workload Identity scenarios. Removed legacy App Engine flexible environment focus.
Major refresh that introduced Cloud Run as a first-class compute option and expanded the observability domain.
PCD (Google Cloud Professional Cloud Developer) is a a challenging, scenario-heavy exam that requires deep hands-on experience and the ability to make architectural trade-off decisions Professional-level exam. Most candidates need 150–300 hours of study spread over 3–6 months for professional and expert-level exams. These exams typically expect prior associate-level proficiency. Most candidates who score consistently above the passing threshold on practice exams pass on their first attempt.
Most candidates need 150–300 hours of study spread over 3–6 months for professional and expert-level exams. These exams typically expect prior associate-level proficiency. 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.
PCD is a recognized credential in the GCP 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 GCP day-to-day or want to move into roles that do.
The passing score for PCD is Not published. The exam contains 50 questions and lasts 2 hr.
The PCD exam fee is $200 USD. Fees are set by GCP and may vary by region; always confirm the current price on the official GCP certification page before booking.
Google Cloud Professional certifications are valid for 2 years. Recertify by re-passing the current version of the exam.
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 PCD. The exam-simulation mode mirrors the real exam: 50 questions in 2 hr, with the same passing threshold of Not published. Browse mode lets you read every Q&A statically.