Why the HMRC CEST Tool Is Flawed and How to Protect Your Status

Check Employment Status for Tax (CEST) is HMRC's free online tool for determining whether an engagement falls inside or outside IR35. It carries an air of authority because HMRC built it, and many end-clients treat a single CEST run as the whole of their reasonable care obligation. That is a mistake. CEST has structural design flaws that have been criticised in tribunal after tribunal, and a determination resting on CEST alone is one of the weakest foundations a status decision can have.

This piece sets out exactly where CEST goes wrong, why HMRC's promise to stand by its output is narrower than it sounds, and how a contractor protects their status when a client leans on the tool. Sister pieces in [the IR35 hub](/guide/ir35-off-payroll-guide/) cover [who carries the fee-payer liability in the supply chain](/blog/fee-payer-responsibilities-liability-supply-chain/) and [how to structure B2B contracts that defend an outside position](/blog/b2b-contracts-defend-outside-ir35/).

What CEST is and what it is not

CEST is a questionnaire. It asks a sequence of yes/no and multiple-choice questions about substitution, control, financial risk, and how the worker is integrated into the engager's business, then returns one of three results: employed for tax purposes (inside IR35), self-employed for tax purposes (outside IR35), or unable to determine. It is an indicative tool, not a statutory determination. The legislation does not require its use, and its output is not binding on a tribunal.

HMRC has stated it will stand behind a CEST result provided the information entered is accurate and the questions are answered honestly, and provided the engagement is operated in line with those answers. That commitment sounds reassuring but it is heavily conditional. If working practices drift from the answers given, or if HMRC later decides an answer was not an accurate reflection of reality, the protection evaporates.

The Mutuality of Obligation gap

The single most serious flaw in CEST is that it does not test Mutuality of Obligation (MOO) in any meaningful way. HMRC's position is that MOO is present in every engagement it is asked to assess, because a contract to do work for payment inherently contains an obligation to do the work and an obligation to pay for it. On that reasoning, HMRC built CEST to assume MOO exists and moved the question off the tool entirely.

The courts disagree with that reading. The case law from Ready Mixed Concrete (1968) onward treats the irreducible minimum of MOO, the ongoing obligation to offer and accept work, as a live test that can point toward self-employment when it is absent. The Court of Appeal and the Supreme Court in HMRC v Professional Game Match Officials Ltd (PGMOL) examined MOO in detail precisely because it matters. By omitting the test, CEST systematically tilts borderline cases toward an inside result.

Over-weighting of control

CEST gives heavy weight to control: who decides how the work is done, where, when, and in what order. Control is a genuine status factor, but it is only one of the three primary tests established in Ready Mixed Concrete alongside personal service and MOO. Modern professional engagements frequently involve a degree of client coordination (security access, site rules, attendance at a kick-off meeting) that CEST can read as control even where the contractor retains genuine autonomy over method and sequence.

The result is that an engagement with strong substitution and no MOO can still be pushed inside by CEST on the strength of routine coordination, because the tool weights control more heavily than the case law does and cannot weigh the absence of MOO against it.

The "undetermined" problem

CEST returns an "unable to determine" result on a meaningful proportion of runs, and on exactly the borderline cases where a determination is most needed. HMRC offers no protection at all for an undetermined output: the end-client is left to make the call itself with no safe harbour. Faced with an undetermined result and an undisclosed appetite for risk, many clients default to an inside determination to remove their own exposure, which transfers the cost to the contractor.

Where the case law has rejected CEST

Tribunals have repeatedly reached status conclusions that diverge from a CEST run, and have been openly critical of the tool's treatment of MOO. The consistent judicial theme is that status is a multi-factorial assessment of the whole picture, painting the realistic working relationship, which a fixed questionnaire cannot capture. A determination defended only by a CEST printout has, in practice, repeatedly failed to satisfy the reasonable care standard required of end-clients under the off-payroll rules.

Why a CEST result does not equal reasonable care

Under the off-payroll working rules in Chapter 10 of ITEPA 2003, the end-client must take reasonable care in reaching its status determination. HMRC guidance is explicit that simply running CEST is not automatically reasonable care, particularly where the inputs are inaccurate, where a blanket approach is applied across a whole contractor population, or where the client ignores facts that contradict the tool. A client that treats one CEST run as the end of the exercise is exposed if the determination is later challenged.

How CEST answers can be manipulated, knowingly or not

The same facts, two different results

CEST output is only as good as the inputs, and the inputs are subjective. The same engagement can return inside or outside depending on who answers and how cautiously. An end-client risk team answering defensively will frame coordination as control and a heavily caveated substitution clause as no real right of substitution. The same facts, answered by reference to the genuine contractual right and the actual working practice, can return outside.

Protecting your status when a client relies on CEST

A contractor cannot force a client to abandon CEST, but can ensure the determination is built on more than a single run. The practical defence is a paper trail that addresses the tests CEST under-weights or omits, gathered before the engagement starts and maintained throughout.

• Obtain an independent contract review that assesses substitution, control, MOO, and financial risk against the case law, not just against CEST.
• Insist the written contract contains a genuine, unfettered right of substitution rather than a "subject to client approval" clause.
• Negotiate explicit no-MOO language covering both during and between engagements, reflecting the post-PGMOL position.
• Keep evidence of working practices that match the contract: emails showing you set your own hours, examples of declining tasks within scope, your own equipment and insurance.
• Ask the client to record the full reasoning behind the determination in the Status Determination Statement, not merely the CEST reference number.
• Where the client relies solely on CEST, raise it directly through the [SDS disagreement process](/blog/status-determination-statement-appeal/) as a reasonable care failure.

Should I run CEST myself before accepting a contract?

Running CEST yourself is a useful diagnostic, but treat it as a temperature check rather than a verdict. If your honest answers return outside, keep the printout and the dated inputs as supporting evidence. If they return inside or undetermined, that is a signal to renegotiate the contract terms or working practices before signing, not a reason to abandon the engagement. The real protection comes from the contract and the working practices behind the answers, not from the tool.

Does HMRC ever audit CEST results?

HMRC does not police every CEST run, but it does open compliance checks into end-clients' off-payroll processes, particularly where blanket inside determinations or sector-wide patterns suggest a lack of case-by-case care. In those checks, a client relying only on CEST and unable to show the underlying working-practice evidence is in a far weaker position than one with a documented multi-factorial assessment. The tool is the start of the reasonable care exercise, never the whole of it.

CEST is a snapshot, not a living record

Even a CEST result that returns outside on accurate inputs only reflects the engagement as it was described on the day it was run. Engagements evolve. A contractor who starts on a defined deliverable can drift into open-ended business-as-usual work, attend the same meetings as permanent staff, and gradually be managed like an employee. None of that is captured by a CEST run completed at the start. HMRC's commitment to stand by the output is expressly conditional on the engagement continuing to match the answers given, so a stale CEST result protects nobody once the reality has moved on.

The practical consequence is that CEST cannot be a one-off box-ticking exercise at onboarding. Where an engagement materially changes scope, the determination needs to be revisited and the working-practice evidence refreshed. A determination that was sound in month one can be undermined by month twelve if no one keeps the picture current, which is another reason the durable protection sits in the contract and the contemporaneous evidence rather than in a single questionnaire result.

The bottom line on CEST

CEST is a free, fast, official-looking tool that omits the test the courts treat as central, over-weights a test the courts treat as one factor among several, and offers no protection at all on the borderline cases where it most often lands. It has a place as an early indicator, but a contractor or client who treats it as the determination is building on sand. The durable protection is a well-drafted contract, working practices that match it, and a determination that reasons through all the tests rather than outsourcing the decision to a questionnaire. For the contractual side of that defence, see [structuring B2B contracts to defend an outside position](/blog/b2b-contracts-defend-outside-ir35/).