← Back to guides

The short version

• ·High indeterminate rate. HMRC's own published statistics show CEST cannot reach a determination in around 22% of cases, leaving contractors and businesses with no protection either way.
• ·No reasoning shown. CEST returns a single word, inside, outside, or undetermined, with no explanation of which factor drove the result.
• ·Mutuality of obligation is barely modelled. Following HMRC v PGMOL (2024), mutual obligations can exist within a single engagement, a nuance CEST's fixed questionnaire does not capture.

What CEST actually does

CEST stands for Check Employment Status for Tax. You answer a series of questions about substitution, control, and the wider arrangement, and it returns one of three results: outside IR35, inside IR35, or unable to determine. That is the whole output. There is no score, no probability, no reasoning, and no reference to the case law that the questions are loosely based on.

Two features of that design cause problems. The first is that the result is binary, so a borderline engagement that a judge might spend days picking apart is flattened into a single word. The second is that CEST takes your answers at face value. It has no mechanism to test whether what you typed reflects how you actually work day to day.

Where CEST and a tribunal part company

A tribunal does not start from a questionnaire. It starts from the three conditions set out in Ready Mixed Concrete (South East) Ltd v Minister of Pensions (1968), which remain the foundation of every status case. There must be personal service, sufficient control, and a wider set of contract terms consistent with employment rather than someone genuinely in business on their own account. The first two are gateway conditions. The third is an overall evaluation, and it is where most modern cases are won or lost. CEST never reaches that evaluation in any meaningful way.

A straightforward example shows why this matters. Take an IT contractor placed via an agency with a software firm, working across multiple client engagements, with a substitution clause in their contract and no obligation on either side to offer or accept further work once a project ends. CEST may return an inside IR35 result or an unable to determine because it struggles with nuanced substitution rights and the overall picture of someone genuinely in business. A tribunal applying Ready Mixed Concrete would start from a very different place: looking at the full pattern of working, not just the answers to a fixed questionnaire.

Three gaps matter most.

It accepts the paper over the practice. In Autoclenz Ltd v Belcher (2011), the Supreme Court held that where written terms do not reflect the true working arrangement, a tribunal looks through the contract to the reality. A polished outside-IR35 contract carries little weight if your working practices contradict it. CEST has no way to detect that divergence, because it only ever sees your answers, not your working life.

Its handling of mutuality of obligation is thin. Mutuality of obligation, the question of whether each side is obliged to offer and accept work, has long been a contractor's defence. The Supreme Court in HMRC v Professional Game Match Officials Ltd (PGMOL) (2024) held that mutual obligations and sufficient control can exist within a single engagement, even where there is no obligation between engagements and even where either side can cancel without penalty. That lowers the bar HMRC has to clear, and CEST does not model the nuance at all. We cover this in detail in our explainer on mutuality of obligation.

It gives you no reasoning to act on. When CEST says inside, it does not tell you which factor tipped the balance, so you cannot tell whether a small change to how you work would change the picture. A determination you cannot interrogate is a determination you cannot use.

Does that mean CEST is wrong?

Not necessarily. Where your contract and your working practices genuinely point to employment, CEST and a tribunal may reach the same place. The point is narrower and more useful than "CEST is wrong". CEST is a blunt instrument applied to text you supply, with a known tendency to return inside or to give up entirely. It tells you what HMRC's questionnaire makes of your answers. It does not tell you what a judge would make of your circumstances, and it cannot show you the financial cost of an inside determination either way.

If uncertainty about your status is the concern, what an HMRC IR35 investigation actually costs explains the scale of that risk.

Frequently asked questions

Is CEST accurate? CEST is consistent rather than accurate. It applies a fixed questionnaire to the answers you provide and returns a binary result. HMRC's own figures show it cannot reach a determination in around 22% of cases, and it does not test whether your answers reflect how you actually work, which is the point on which tribunals decide most cases.

Is an HMRC CEST result legally binding? HMRC has said it will stand by a CEST result provided the answers were accurate and given in good faith. That protection falls away if your answers do not match reality, which is exactly the gap the Autoclenz principle addresses.

Why does CEST so often say inside IR35 or fail to decide? CEST's questions struggle with the subtler factors that favour genuine contractors, particularly mutuality of obligation and the overall in business on own account picture. When those points are finely balanced, the tool tends toward an inside result or an unable to determine outcome.

What is a better alternative to CEST? Any approach that weighs the same factors a tribunal would and tests your contract against your working practices. Our contract checker returns a probability based result with reasoning grounded in UK case law, rather than a single word with no explanation. Unlike CEST, the checker shows its working: you get a probability-based result with the case law reasoning behind each factor, not a single word with no explanation. For a full comparison of the options, see our full guide to genuine CEST alternatives.