“"Excellent question coverage and very easy to practice. It helped me stay consistent throughout my preparation."”
Thomas
Treasury Analyst
The Certified Treasury Professional (CTP) exam runs 170 multiple-choice questions inside a four-hour testing appointment — three and a half hours to actually answer, with the rest set aside for an optional tutorial and a closing survey. Of those 170 questions, 150 count toward your score and 20 are unscored pretest items mixed in without any way to tell them apart from the scored ones. That matters for how you pace yourself: treat every question as if it counts, because you can't identify the ones that don't.
AFP has also been explicit about question difficulty in its published specifications: roughly a third test basic facts and terms, another 38–42% test understanding of concepts, 23–27% require applying those concepts to a specific business situation, and 10–15% involve actual math. That last group trips up more candidates than the percentage suggests, since AFP members often haven't done a WACC or NPV calculation by hand since business school.
AFP's current test specifications (the version governing the 2026–2028 testing cycle) split the exam across five domains. The ranges below are published by AFP; the weight percentages are calculated from those ranges against the 170-question exam.
Domain | Focus | Questions | Approx. Weight |
|---|---|---|---|
Domain 1 | Maintain corporate liquidity (cash positioning, forecasting, banking structure, FX, trade and intercompany financing) | 55–60 | ~32–35% |
Domain 2 | Manage capital structure and long-term capital (debt/equity issuance, syndicated agreements, M&A impact, hedging) | 33–37 | ~19–22% |
Domain 3 | Manage internal and external relationships (banking relationships, service provider selection, internal stakeholders) | 12–17 | ~7–10% |
Domain 4 | Monitor and control corporate risk exposure (policy compliance, fraud detection, counterparty risk, regulatory reporting) | 23–27 | ~14–16% |
Domain 5 | Assess the impact of technology on treasury (ERP systems, cybersecurity, payment fraud, AI-related risk) | 23–27 | ~14–16% |
Domain 1 alone accounts for roughly a third of the exam, which is why cash positioning and forecasting deserve more repetition in practice than any other topic.
Domain 5 catches more experienced treasury professionals off guard than any other section. It's the newest domain in the specifications, it covers ground — AI-related risk, payment card fraud, cyber and email-scam exposure — that older reference material doesn't treat in depth, and candidates who've spent a decade doing cash positioning by feel often haven't kept pace with how treasury technology risk is framed on paper.
Domain 3 is the other soft spot, not because the content is hard but because it's judgment-heavy. Questions about selecting a banking partner or managing an internal stakeholder relationship don't have a clean formula behind them, so candidates who study by memorizing definitions get caught by questions that are really asking "what would you actually do here."
And the math-based subset (10–15% of the exam) is underestimated simply because most CTP candidates come from a treasury operations background rather than a pure finance-analyst one — WACC, effective interest cost, and DSO/DPO calculations show up, and rustiness there costs points that have nothing to do with treasury judgment.
Because roughly a quarter of the exam tests applied judgment rather than recall, drilling flashcard-style definitions only prepares you for part of the test. Practice sessions built around scenario questions — a treasury team deciding between two banking structures, or evaluating whether a counterparty exposure needs hedging — do more to build exam-day readiness than repeating terminology.
Since scoring isn't penalized for wrong answers, it's also worth practicing the habit of never leaving a question blank, even under time pressure, and using full-length timed sessions to get a feel for the roughly 1.2 minutes per question you'll have on test day If you're preparing for Certified Valuation Analyst (CVA), try our online practice test.
Recent testing windows have produced pass rates in the 43–51% range, so a meaningful share of first-time candidates don't clear the bar. The recurring patterns behind that:
Treating the exam as a knowledge test rather than an applied-judgment test, especially in Domains 1 and 4
Under-preparing Domain 5 because it feels tangential to "real" treasury work
Skipping calculation practice and losing points on the 10–15% of questions that require math
Running out of time in the back third of the exam after over-deliberating early questions
Studying Domain 1 and 2 heavily because they're the biggest, while neglecting Domain 3's relationship-management scenarios entirely
New to treasury or early-career: Budget close to the full 6–12 month window AFP references for exam preparation. Build a domain-by-domain foundation first, then layer in scenario practice once the vocabulary is solid.
Working treasury professional with hands-on experience: Domain 1 and parts of Domain 4 will likely feel familiar from daily work. Concentrate preparation time on Domain 5's technology and risk content and on Domain 2's capital structure calculations, which are less likely to come up in a typical operating role.
Retaking the exam: Since AFP doesn't release a raw score or a list of missed questions, use domain-level practice results as your best proxy for where you lost points. A near-miss score usually reflects several scattered misses rather than one weak domain, so mixed-topic practice sessions tend to help more than re-drilling a single area.
AFP updates the CTP test specifications on a three-year cycle tied to a new edition of its Essentials of Treasury Management reference text — the current specifications govern the 2026–2028 testing cycle. Domain 5 has been the most active area of change in recent cycles as AFP has expanded its treatment of cybersecurity, payment fraud, and AI-related risk within treasury technology. Candidates studying from older reference material should specifically double-check that their Domain 5 content reflects the current specifications rather than an earlier edition.
Last updated on Oct, 9 2026