- CT-MBT (Model-Based Testing) is the ISTQB specialist certification that teaches you to design tests from models (workflows, state diagrams, decision tables) and to generate test cases systematically and, often, automatically.
- Current version: syllabus v1.1, released February 23, 2024 (it replaces v1.0 from 2015).
- Prerequisite: the ISTQB Foundation Level (CTFL) certificate. At least six months of development or testing experience is recommended.
- Exam: 40 questions, 40 points, pass mark 26 (65%), 60 minutes (75 if it isn't in your native language).
- 5 chapters and a minimum of 725 minutes (12 h 05 min) of accredited training; the heaviest chapter is modeling (250 min).
- Useful for testers, analysts, automation engineers, product owners and developers dealing with complex requirements, regulated systems or heavy regression.
- The exam isn't included in any course: buy it from iSQI (20% coupon for Cool Testers students) or through us if you need an invoice.
Have you ever had 300 hand-written test cases, a business rule changes, and you have to check them one by one to see which are affected? Model-Based Testing (MBT) tackles exactly that: instead of maintaining hundreds of loose test cases, you maintain one model and generate the cases from it. ISTQB CT-MBT gives you the framework to do it well. This guide explains what the v1.1 syllabus covers, what the exam looks like and how to prepare.
For the big picture of the certification scheme, see our main ISTQB certification guide.
What CT-MBT is and where it sits in the ISTQB portfolio
The syllabus describes MBT as an approach in which you design a comprehensive model (usually with tools) based on the project's test objectives, and that model serves as a test design specification with enough formal detail to generate test cases automatically.
Put simply: you draw how the system behaves (or how it's used, or how you want to test it), apply a selection criterion ("cover all transitions", "cover all high-priority requirements") and get your test cases. When something changes, you update the model and regenerate.
Current version
The current version is v1.1 (February 23, 2024). Compared with v1.0 (2015), content was reorganized, more advanced material was removed and the appendix with the simple modeling language was changed. If you come across older study material, check that it matches v1.1. At the time of writing, the iSQI exam page still mentions v1.0, so confirm the version when you book.
Where it fits
CT-MBT belongs to the specialist stream of the ISTQB portfolio and is taken after ISTQB Foundation Level (CTFL 4.0). The syllabus suggests the Advanced Level Test Analyst, Technical Test Analyst and Test Management certifications as next steps, followed by Expert Level.
Who it's for and prerequisites
According to the syllabus, CT-MBT is aimed at anyone involved in software testing: testers, test analysts, test engineers, consultants, test managers, user acceptance testers and developers. It also suits people who need a basic understanding of testing: project managers, quality managers, architects, development managers and business analysts.
In practice, you'll get the most out of it if you work with:
- Complex business rules (insurance, banking, payroll, tax calculation).
- Systems with many states (approval workflows, devices, control systems).
- Regulated or critical domains requiring traceability between requirements, code and tests.
- Heavy regression and frequently changing requirements.
Exam entry requirements
- Mandatory: the ISTQB Certified Tester Foundation Level certificate.
- Recommended: at least six months of experience as a system or acceptance tester, or as a developer.
- Recommended: a course accredited by an ISTQB Member Board.
What you'll be able to do: business outcomes
The syllabus defines five Business Outcomes. A Model-Based Testing Certified Tester can:
| Outcome | What it means in practice |
|---|---|
| MBT-BO1. Collaborate in an MBT team using standard terminology and established MBT concepts, processes and techniques | Speak the same language as analysts, automation engineers and MBT tools. |
| MBT-BO2. Apply and integrate MBT in a test process | Know which test activities change and how MBT fits sequential or agile lifecycles. |
| MBT-BO3. Effectively create and maintain MBT models | Model at the right level of abstraction, following good practices. |
| MBT-BO4. Select, create and maintain test work products from MBT models considering risk and value | Choose selection criteria that give the coverage you need without test case explosion. |
| MBT-BO5. Support the organization in improving its QA process | Assess ROI and introduce MBT gradually and measurably. |
Syllabus content chapter by chapter
The v1.1 syllabus has five examinable chapters and requires at least 725 minutes of training in accredited courses. Learning objectives range from K1 (remember) to K3 (apply).
| Chapter | Minimum time | Key topics |
|---|---|---|
| 1. Introduction to Model-Based Testing | 90 min (K2) | Benefits, misleading expectations, MBT activities and work products, lifecycle integration, support for requirements engineering |
| 2. MBT Modeling | 250 min (K3) | Workflow and state models, model focus (system, environment, test), modeling languages, model quality, guidelines, reuse, tools |
| 3. Selection Criteria for Test Case Generation | 205 min (K3) | Criteria families: model, requirements and data coverage, random, scenario-based and project-driven; degree of automation; good practices |
| 4. MBT Test Implementation and Execution | 120 min (K3) | High- vs. low-level test cases, manual and automated execution, offline and online MBT, impact of changes, test adaptation layer |
| 5. Evaluating and Deploying an MBT Approach | 60 min (K2) | ROI factors, objectives vs. approach characteristics, metrics and KPIs, good practices, cost factors, tool integration |
Chapter 1. Introduction to Model-Based Testing (90 min)
It starts with the "why". The syllabus groups the motivations into two areas:
- Effectiveness: modeling fosters communication with stakeholders, helps detect misunderstandings in requirements, makes it easier to involve business analysts when models are graphical, and lets you generate and analyze test cases before the system exists.
- Efficiency: early model verification prevents defects, models can be reused, different test suites can be generated from the same model with different criteria, and the model becomes a single point of maintenance when requirements change.
Just as important are the misleading expectations. The syllabus warns against believing that "MBT solves all problems", that "MBT is just a matter of tooling" (buying a tool shouldn't be the first step), that "models are always correct" (they contain defects too and must be reviewed), and against test case explosion when combinatorial generation is used without control.
You'll also see what changes in each test activity (planning, analysis and design, implementation and execution, monitoring and control, completion), which work products go into and come out of MBT, and how it fits sequential and agile lifecycles: models that grow each iteration, user stories linked to the model for bidirectional traceability, and MBT driving ATDD.
Example: in a Scrum team, the tester models the "request a refund" flow during refinement. Reviewing it with the product owner reveals a path nobody had defined: what happens if the order was already partially refunded? The defect was found in the requirement, before a single line of code was written.
Chapter 2. MBT modeling (250 min)
The heart of the certification and the chapter with the most time. There are two key K3 objectives: develop a simple model (fewer than 15 elements) using a workflow-based language and another using a state transition-based language. The syllabus appendix defines a simple graphical notation for both.
Other topics in the chapter:
- Questions to ask when modeling: which quality characteristics to model (usually functionality, but also performance or security), which language to use and what level of abstraction. More abstraction simplifies the model but makes test adaptation for execution harder.
- Model focus: structural (e.g., classification trees) or behavioral (activity diagrams, BPMN, state diagrams); and system, environment/usage (e.g., Markov chains describing expected usage) or test models.
- Models depend on test objectives: the model for testing business rules is not the one for testing an app's navigation.
- Languages: for structural, data and behavioral models, plus integrated languages such as UML. Good fits include state diagrams for control systems, activity diagrams or BPMN for end-to-end testing of information systems, and decision tables for business rules.
- Model quality: syntactic (correct per the language rules), semantic (valid with respect to what it describes) and pragmatic (suitable for the objective and the generator).
- Typical mistakes: too much or too little detail, and trying to build one model that covers everything.
- Linking requirements and risks to the model: lets you generate tests per requirement, measure requirements coverage, analyze the impact of changes and get the traceability matrix automatically.
- Modeling guidelines, when reusing design or requirements models is (and isn't) appropriate, tool types, and iterative model development with review and validation.
Practical example: you model a user account with the states Registered → Active → Locked → Active and Closed. Transitions are "verify email", "3 failed attempts", "unlock by support" and "close account". With that diagram of fewer than 15 elements you can already generate tests covering every state and every transition, including the one people almost always forget: trying to close a locked account.
Chapter 3. Selection criteria for test case generation (205 min)
A model can produce thousands of paths. Test selection criteria decide which ones become test cases. The syllabus groups them into families:
- Coverage of requirements linked to the model.
- Coverage of model elements: states, transitions and decisions in state diagrams; activities and gateways in business process models; conditions and actions in decision tables; statements and conditions in textual models.
- Data-related criteria: equivalence partitioning and boundary value analysis.
- Random, scenario- or pattern-based and project-driven criteria (e.g., risk or priority).
You must be able to generate test cases from a model to meet a given objective (K3) and apply given criteria to a model (K3), and recognize how these criteria relate to Foundation Level test techniques. You'll also cover the three degrees of automation (manual, semi-automated and automated generation), the pros and cons of each criterion (random criteria, for example, can generate long tests with no business meaning) and how to combine them: by composition (intersection: cases meeting all criteria) or by addition (union: cases meeting at least one).
Example: in the user account model, "all states coverage" might give you just 2 cases; "all transitions coverage" gives you more; and adding "boundary values for failed attempts" (2, 3 and 4) combines a model criterion with a data criterion.
Chapter 4. MBT test implementation and execution (120 min)
This is where the model becomes executable tests. Key topics:
- High-level vs. low-level test cases: high-level cases have no concrete values (useful for review with analysts); low-level cases have concrete data and expected results (ready to execute).
- Manual or automated execution. For automation, cases are sometimes converted straight into scripts; other times a test adaptation layer (similar to keyword-driven testing) separates test design from implementation details.
- Offline MBT (scripts and expected results are generated first and executed afterwards) vs. online or "on-the-fly" MBT (each step is generated after executing the previous one, and the result can change the path taken).
- Impact of changes (K3): what to do when a requirement, only the interface, or the test objective changes.
- Test adaptation activities: exporting cases as scripts, implementing keywords in the execution tool, linking abstract data to concrete data and handling preconditions between scripts.
Example: the UI changes and the "Unlock" button becomes a menu item. Behavior is the same, so you don't touch the model: you just update the "unlock account" keyword in the adaptation layer and every generated test keeps working. For the automation side, see our guide to Test Automation Engineering (CTAL-TAE).
Chapter 5. Evaluating and deploying an MBT approach (60 min)
The last chapter looks at MBT as an investment. It covers:
- A five-step adoption path: awareness, interest, evaluation, trial use (with defined KPIs) and adoption.
- ROI factors: introduction and recurring costs versus savings such as early requirements validation and reuse of redundancy-free test design work products.
- How project objectives relate to the characteristics of the MBT approach, and which metrics and KPIs measure progress and results.
- Good practices for test management, change management and collaboration, cost factors, and the need to integrate the MBT tool with configuration, requirements and test management and with test automation tools.
Example: before buying licenses, you pick a module with heavy regression, define KPIs (design hours per release, defects found in requirements, update time after a change) and run a two- or three-sprint pilot. The data tells you whether to scale.
Download the official syllabus
The current syllabus (v1.1) is in English. Download it here and check the ISTQB page for sample exams:
The CT-MBT exam
Verified structure
| Item | Value |
|---|---|
| Questions | 40, multiple choice |
| Total points | 40 |
| Pass mark | 26 points (65%) |
| Duration | 60 minutes; 75 minutes if it isn't in your native language (+25%) |
| Prerequisite | ISTQB Foundation Level |
Source: ISTQB Exam Structure and Rules Tables v1.18 (May 2026). Based on study time, chapters 2 and 3 carry the most weight; prioritize modeling and applying selection criteria.
Exam formats
- Online with remote proctoring (iSQI FLEX), 24/7: book at least 2 hours ahead. You need a webcam, a microphone, a second camera (your phone) and a government-issued ID.
- Pearson VUE: at a test center or online with OnVUE.
- On-site at your company for groups of 6 or more: Cool Testers brings a proctor.
Paper-based exams are no longer offered. Language: at the time of writing, iSQI offers this exam in English and Brazilian Portuguese. If neither is your native language, you get 25% extra time. The certification has the same value in any language.
Where to buy it
- Directly at iSQI.org, our partner and official exam provider, with a 20% coupon off the price on their site for the first attempt for Cool Testers students. Recommended if you're paying yourself.
- Through Cool Testers (certification exam, 1 or 2 attempts) if your company pays or you need an invoice. You have 1 year to take it. Contact us to confirm availability.
Costs in your currency: main ISTQB certification guide.
How to prepare for CT-MBT
- Refresh Foundation Level techniques. State transitions, decision tables, partitions and boundary values are the basis of the selection criteria.
- Learn the appendix notation. The K3 objectives ask you to build workflow and state models with fewer than 15 elements using that simple notation.
- Model three systems you know. A login, a shopping cart and an approval workflow. For each, generate cases with two different criteria and compare.
- Practice change impact. Change a rule and decide what gets updated: the model, the criterion or the adaptation layer?
- Memorize the K1 lists: MBT work products, language categories, model quality characteristics, tool types, cost factors.
- Take the official sample exam and review the rationale for each answer.
- If you can, try an MBT tool or even a diagram editor: modeling hands-on prepares you better than reading.
Benefits and career paths
- Better test design. Even without an MBT tool, modeling forces you to see paths, states and rules that hand-written cases tend to miss.
- You become the bridge between business and development. A graphical model is much easier to review with an analyst than 200 cases in a spreadsheet.
- You bring traceability. Highly valued in banking, insurance, healthcare, automotive and other regulated sectors.
- It complements automation. MBT plus a keyword-driven adaptation layer is a very maintainable way to automate large suites.
- Roles where it adds value: test analyst, QA engineer on complex systems, automation engineer, quality consultant, requirements analyst with a testing focus.
We don't quote salary figures because they vary widely by country and company; the certification adds most value when you can back it up with models you've built on real projects.
Related certifications and next steps
| Certification | Relationship to CT-MBT |
|---|---|
| CTFL 4.0 | Entry requirement; provides the black-box techniques |
| CT-AcT Acceptance Testing | Uses BPMN and DMN models to derive acceptance tests; shares the idea of coverage on models |
| CTAL-TA Test Analyst | Natural next step to deepen test design techniques |
| CTAL-TTA | Goes deeper into technical and non-functional aspects |
| CTAL-TAE and CT-TAS | To turn generated tests into a solid automation architecture |
| CTAL-TM | For whoever will decide on and measure MBT adoption |
Frequently asked questions
What is Model-Based Testing in simple terms?
It's designing tests from a model of the system, its usage or the test itself (for example, a state diagram or a workflow) and generating test cases by applying selection criteria, often with the help of a tool.
What is the current CT-MBT syllabus version?
v1.1, released on February 23, 2024. It replaces v1.0 from 2015.
How many questions are on the CT-MBT exam?
40 questions worth one point each. You pass with 26 (65%). You have 60 minutes, or 75 if it isn't in your native language.
Do I need to know how to code for CT-MBT?
No. The exam covers modeling, selection criteria and deploying the approach. Automation knowledge helps you understand the adaptation layer, but it isn't required.
What do I need to take the exam?
The ISTQB Foundation Level certificate. ISTQB also recommends at least six months of development or testing experience.
Can I take CT-MBT in Spanish?
At the time of writing, iSQI offers it in English and Brazilian Portuguese. If neither is your native language, you get 25% extra time. The certification has the same value in any language.
Do I need to buy an MBT tool?
Not to get certified. The syllabus itself warns that acquiring a tool shouldn't be the first step: first define measurable objectives and run a pilot.