What evidence does CQC look for?
Reviewed September 2026 ยท Written by Complys and checked against CQC's published guidance (sources below).
CQC assessors do not want to be told a service is good โ they want to be shown. This guide is a practical reference to what "shown" means: the specific evidence assessors look for under each of the five key questions, and, just as importantly, what separates evidence that helps you from evidence that does not. If you know exactly what proof each question needs and you can put your hands on it quickly, an assessment stops being a test you might fail and becomes a description of a service you already run well.
Two kinds of evidence: records and reality
Everything below falls into one of two kinds. There are records โ the certificates, training logs, care plans and audits that document how the service runs over time. And there is reality โ what an assessor sees when they walk through the service and what people tell them when they talk. The two have to agree. Records exist to evidence the things a single visit cannot capture; observation and conversation test whether those records are true. A service whose paperwork says one thing while its corridors say another has a bigger problem than a missing certificate, because it suggests the documents are for show. Keep that in mind as you read: the goal is not a full folder, it is a folder that honestly describes a good service.
Evidence for Safe
Safe is the most evidence-heavy question because so much of it is documentable. Assessors typically look for:
- Current building safety certificates โ fire risk assessment, fire drills and alarm tests, legionella risk assessment and water temperature monitoring, gas safety, fixed electrical (EICR), and LOLER inspections for hoists and lifting equipment.
- Safe medicines management โ recent medicines audits, clear administration records (MAR charts), safe storage and stock control, and evidence that errors are identified and acted on.
- Individual risk assessments that are current and genuinely reflect each person's needs, not a generic template.
- Safeguarding records โ concerns, referrals and outcomes โ showing the service recognises and reports abuse and neglect appropriately.
- Safe recruitment evidence in staff files โ DBS checks, right to work, references and interview records.
- Accident and incident logs with evidence that lessons are learned and practice changes as a result.
- Staffing records that show the right numbers and skill mix on every shift.
Evidence for Effective
Effective is about whether care works and follows current best practice. Look to have ready:
- Needs assessments and care plans that follow recognised good practice and are reviewed as needs change.
- Up-to-date mandatory training for every member of staff, with no quiet expiries hiding in the matrix.
- Supervision and appraisal records that show staff are developed and their competence is checked, not just that a course was booked.
- Mental Capacity Act evidence โ capacity assessments where needed, recorded best-interests decisions, and lawful handling of any deprivation of liberty.
- Nutrition and hydration records, and evidence of supporting people's wider health โ access to GPs, dentists and other services.
- Evidence of working with other services so that care joins up rather than fragmenting at handovers.
Evidence for Caring
Caring is weighted towards observation, but records still support it:
- Care plans that show people โ and, where appropriate, their families โ are genuinely involved in decisions about their care.
- Evidence that dignity, privacy and confidentiality are protected in everyday practice.
- Feedback from people who use the service and their relatives, including compliments and how everyday concerns are handled.
- Evidence of support for emotional wellbeing and for maintaining relationships that matter to people, not only physical care needs.
What assessors observe here matters as much as any record: how staff speak to people, whether people are rushed, and whether the service feels like a home or a process. No document offsets a visit that feels institutional.
Evidence for Responsive
Responsive is about care organised around the person. Expect assessors to look for:
- Person-centred care plans that reflect individual preferences, culture, routines and communication needs โ and that are reviewed as things change.
- Evidence of meeting the Accessible Information Standard for people with communication or sensory needs.
- A complaints log that is genuinely used โ concerns recorded, investigated, resolved and learned from โ rather than conspicuously empty.
- Evidence of activities and meaningful occupation suited to the people using the service.
- Where it applies, end-of-life care planning that respects people's wishes.
Evidence for Well-led
Well-led ties the others together, and its evidence is about governance and leadership:
- An audit schedule with completed audits โ medicines, care plans, infection control, health and safety โ and, crucially, evidence that findings become dated actions that get closed.
- A risk register or equivalent showing the service knows its own risks and is managing them.
- Your record of statutory notifications to CQC, evidencing the events that required them.
- Evidence that duty of candour is understood and applied when things go wrong.
- Records of engaging staff and people who use the service โ meetings, surveys and what changed as a result.
- A clear improvement plan that shows the service knows where it is heading and how it will get there.
The single strongest signal under Well-led is not a perfect record but a service that visibly finds its own problems and fixes them. Leaders who can point to what they got wrong and what they did about it fare better than those who claim nothing ever goes wrong.
What makes evidence strong โ or weak
The same document can be strong evidence or worthless depending on three things. Currency: is it in date? An expired certificate or a lapsed training record is worse than none, because it shows the system that should keep it current has failed. Specificity: does it describe this person, this service, this month โ or is it a generic template that could belong to anyone? A care plan that names real preferences and needs is evidence; one that reads the same for every resident is not. Closure: does it finish the loop? An audit that finds a problem is only half an audit; the evidence assessors value is the action that followed and the check that it worked. Hold every piece of evidence up against those three tests and you will quickly see which parts of your service are genuinely well-run and which just look tidy.
The difference between having evidence and producing it
There is a gap that catches out even good services: the difference between having evidence somewhere and being able to produce it on request. When an assessor asks for one named person's current risk assessment, or one member of staff's full training history, or last quarter's medicines audit and the actions from it, the clock is effectively running. Evidence spread across filing cabinets, spreadsheets and individual inboxes may all exist and still fail you, because it cannot be assembled in time. If retrieving a specific record takes longer than a few minutes, that is a finding in itself โ and it is the easiest kind of weakness to fix before an assessment ever happens.
Have every piece of evidence at your fingertips
Complys organises your evidence around the five key questions and produces an evidence pack on request, so nothing is ever lost in a filing cabinet. It helps you organise and evidence your compliance โ it does not certify your service or submit your statutory notifications for you; those stay with you as the provider.
Sources
- CQC โ Assessment framework
- CQC โ The 5 key questions we ask
- CQC โ Assessing quality and performance
This guide is general information, not regulatory advice. CQC's assessment approach continues to evolve โ always check CQC's own current guidance for your service type.
What CQC looks for โ FAQs
What evidence does CQC look for?
Evidence that your service is safe, effective, caring, responsive and well-led โ in the form of records people can inspect and things assessors can observe. That spans building safety certificates, training records, staff files, medicines audits, care plans, complaints and incident logs, governance audits and your record of statutory notifications, alongside what people who use the service and staff actually say.
Is it enough to simply have the evidence?
No. The test is whether you can produce the right evidence quickly and whether it reflects reality. Evidence that takes an hour to find, or that describes a service that no longer matches what an assessor observes, does you little good. Being able to produce a person's current risk assessment or a member of staff's full training record within minutes is itself a sign of a well-run service.
What makes evidence strong rather than weak?
Strong evidence is current, specific, and closes the loop โ an audit that leads to a dated action that is then completed; an incident that is recorded, investigated and used to change practice. Weak evidence is out of date, generic, or stops halfway โ a policy nobody has acknowledged, an audit with no follow-up, a care plan written once and never reviewed.
Does CQC only look at documents?
No. A large part of the picture comes from observation and conversation โ how staff interact with people, whether the environment is calm and dignified, and what people who use the service, families and staff say. Documents evidence the things that cannot be seen in a single visit; observation tests whether the documents are true.
What is the most common evidence gap?
Lapsed mandatory training and expired building safety certificates are the most common avoidable gaps, followed closely by audits that are carried out but never acted on. All three are easy to fix in advance and easy for an assessor to spot.
How should evidence be organised?
So that it maps to the five key questions and can be retrieved on request. However you keep it, the practical test is the same: if an assessor asks for a specific record, can you find it quickly and is it current? A live system that flags what is drifting out of date turns this from a periodic scramble into an everyday state.