Last updated July 29, 2026
The ISTQB Certified Tester Advanced Level Agile Tester (CTAL-AT) v2.0 exam gives you 40 questions worth 52 points in total, and you need 34 points to pass — 65.4%. You get 90 minutes, or 113 minutes if you sit the exam in a language that is not your native language. The only formal prerequisite is a valid ISTQB Foundation Level certificate. The syllabus was released by the ISTQB General Assembly on 17 April 2026, runs to 70 pages, and covers six examinable chapters.
CTAL-AT v2.0 is not a refresh of the old Agile Extension. It is a new Advanced Level certification that replaces two retiring ones at once: the Foundation Level Agile Tester (CTFL-AT) and the Agile Technical Tester (CT-ATT). Both are flagged as retiring, with a sunset date of 6 May 2027 for English exams and 6 November 2027 for other languages.
There is no grandfathering. Holding CTFL-AT or CT-ATT does not convert into CTAL-AT v2.0, and neither of them counts as a prerequisite either. If you already hold CTFL-AT, the honest answer is that you sit a new and harder exam: Advanced Level cognitive expectations, including K4 Analyze objectives, which CTFL-AT never contained.
The single most useful fact about this exam is that the questions are not worth the same. A K2 question is worth 1 point, a K3 question 2 points, and a K4 question 3 points. The paper contains 31 K2, 6 K3 and 3 K4 questions: 31 + 12 + 9 = 52. There are no K1 questions at all — the syllabus says in plain text that K1 learning objectives are not present in it.
Do the arithmetic and the consequence is uncomfortable. Nine questions out of forty — 22.5% of the paper — carry 21 of the 52 points, roughly 40% of your score. Your whole margin for error is 18 points (52 − 34). Fail all three K4 questions and you have spent half of that margin on three questions, before making a single mistake anywhere else. That is why you cannot prepare by counting questions per chapter. Prepare by points.
Chapter 4 (Shift Left) and Chapter 5 (Agile Approaches and Test Techniques) together account for 21 of the 40 questions and 29 of the 52 points — more than half of the exam between them. Chapter 6 (Test Automation and Test Tools) is worth two questions. Chapter 3 is the priciest per question: it holds two of the three K4 questions.
The official study times point the same way. The syllabus prescribes 60 / 60 / 210 / 135 / 285 / 30 minutes for chapters 1 to 6 — 780 minutes, 13 hours in total, divided over at least two days. Chapter 5 alone takes more than a third of that time. Where ISTQB spends its own training minutes is the most reliable public signal of priority you will get.
The three K4 questions sit on exactly three learning objectives. Knowing which ones is the highest-leverage thing on this page.
3.1.2 — outline a project test strategy using testing quadrants. You have to produce a strategy, not describe the model: Q1 to Q4 chosen against context factors such as domain complexity and business criticality, delivery frequency, observability, tooling and infrastructure maturity, and risk appetite.
3.4.1 — select appropriate test process improvement measures based on metrics. You are handed a metric signal and must pick the measure that answers it, from root cause analysis and session-based test management to service virtualization, whole-team defect triage and tracking test debt.
5.1.4 — analyze user stories and epics to create test charters. A charter needs at minimum a mission or test objectives plus a scope; the 5W1H technique, also known as the Kipling method, derives the rest. Acceptance criteria give you lightweight test oracle notes, not exhaustive oracles.
The six K3 questions sit on three objectives, two questions each.
4.1.5 — apply user story slicing to reach testable stories, across the six dimensions: workflow, data complexity, interface, scenario, vertical slicing and constraints. Rewrite acceptance criteria for each slice instead of copying the originals, and make the slices together cover the whole story.
5.1.5 — apply exploratory testing in Agile software development: timeboxed sessions of 60 to 120 minutes, with pre-session setup, a test execution phase and a post-session review that compares charter against reality.
5.3.1 — use test smells to evaluate test case quality, across the seven categories and nineteen named smells the syllabus lists for manual test cases, each with its own remedy.
Open with Chapter 5 and its 285 minutes. It is 18 points on its own, and it carries the newest material: exploratory sessions, heuristics, the five test mnemonics, test tours, mob testing, pair testing, vibe testing and test smells.
Then Chapter 3, 210 minutes for 11 points, including two K4 questions. Practise writing a project test strategy and matching improvement measures to metrics rather than memorising definitions.
Then Chapter 4, 135 minutes for 11 points. Drill slicing on real stories, and learn example mapping, storyboarding and testboarding, testware as requirements, and the three named biases.
Chapters 1 and 2 are 60 minutes each and 10 points of pure K2. They reward flat recall and comparison: test types in and after an iteration, end-to-end testing, formal versus holistic testing, regression approaches, the whole team approach and tissue testers.
Chapter 6 comes last, 30 minutes for 2 points. Do not over-invest here — two K2 questions is the entire return.
Finish with a timed full mock, scored in points rather than questions, and spend your remaining hours where you dropped K3 and K4 points.
Plenty of Agile testing courses and blog posts teach material that is simply not in this syllabus, and studying it costs you time you do not have. A full-text check of the 70-page syllabus finds no occurrences of INVEST, the 3C model, story points, planning poker, velocity as a planning metric, three amigos, sprint zero, Definition of Ready, MoSCoW, impact mapping, shift right, mutation testing, or any specific tool name — Chapter 6 discusses tool categories only. Gherkin and given/when/then never appear: BDD is discussed, its syntax is not. There is no technical debt in the syllabus either, only test debt in 3.4.1.
The release notes also move a block of familiar topics out to CTFL v4.0, where they are no longer examined here: the test-first approach, the test pyramid, independent testing, collaborative user story creation, retrospectives, and test effort estimation. Testing quadrants survive, but only through objective 3.1.2.
ISTQB publishes an official v2.0 sample exam, and it is unusually useful here because it is a structural copy of the real paper: 40 questions for 52 points. Treat it as your calibration run rather than as your practice volume — you only get one honest first attempt at it, so save it until you have worked through the heavy chapters.
Sit mock-1 cold, under exam conditions: 90 minutes on a timer, no syllabus, no notes. Score yourself in points out of 52 against the 34-point bar, and check the K3 and K4 items separately — that is where the exam is won or lost. Then read every rationale, including those for questions you got right.
Space the remaining mocks out. Take mock-2 after you have reworked your weak chapters, and keep mock-3 for the last few days before the real thing. Short factual questions are answered on the CTAL-AT FAQ page, and the chapter-by-chapter breakdown lives on the certification overview.
One thing we want to be explicit about: our questions are original, written by hand against the published syllabus and its learning objectives. They are not real exam questions and not dumps — nobody reproducing live exam content does it legally or accurately. They exist to make you reason at the level the exam demands.
As of today we have three full mock exams for CTAL-AT — mock-1, mock-2 and mock-3 — drawn from a pool of 124 questions, with all six chapters covered and a chapter page open for each of them. Every mock uses exactly the official structure: 5 / 5 / 7 / 9 / 12 / 2 questions across the six chapters, a 31 K2 + 6 K3 + 3 K4 split, and the same point weights as the real exam — 1, 2 and 3 points, 52 in total. Because both the structure and the weights match, your percentage here is directly comparable to the percentage you will see on exam day.
The scenarios differ deliberately. Mock-1 leans on fintech, delivery and booking systems; mock-2 on telemedicine, online education and logistics; mock-3 on legacy cores, distributed teams and multi-channel services.