What Ofsted inspectors ask to see in a children's home

Inspectors do not read everything. They case-track a small number of children in depth and follow the thread wherever it goes, which means a handful of records get read very carefully indeed. Knowing which ones, and what is being tested in each, is most of what preparation actually is.

By Mason Hughes, 7 years in children's residential care · Updated 30 July 2026

How an inspector actually works

The mental model that helps most is this: an inspector is not auditing your filing. They are testing whether four things tell the same story. What the records say, what the staff say, what leaders say, and what the children’s lives actually look like. Where those four agree, the visit is calm. Where they diverge, everything else becomes about the divergence.

That is why case tracking is the core method. An inspector picks two or three children and follows them all the way through: the referral, the placement plan, the risk assessment, the daily logs, the incidents, the key-work, the education, the health appointments, the complaints, the missing episodes. One child, followed properly, exposes more than a hundred skim-read files.

The records they ask for, and what each one is testing

Daily logs

Read in runs, not in isolation. The test is whether a stranger could reconstruct a child’s week and understand what the home was worried about. Vague, cheerful entries across a fortnight in which an incident occurred are a much bigger problem than a blunt entry describing something that went badly. There is a longer treatment in the daily logs guide.

Incident and physical intervention records

The test is the trail, not the event. Was de-escalation attempted and recorded? Is the child’s own account there? Was there a debrief, and did anything change afterwards? A restraint record that ends at the moment the hold finished is an incomplete record.

Safeguarding concerns and referrals

Followed from first concern to outcome. Inspectors look for the gap between a concern being noticed and it being acted on, and for concerns that were recorded and then quietly went nowhere. This is the area where a limiting judgement most often originates.

Care plans, placement plans and risk assessments

Checked for currency and for whether they match reality. A risk assessment that has not been updated since a significant incident is read as evidence that the paperwork and the practice have separated.

Key-work and children’s views

The test is whether children’s wishes and feelings genuinely shape decisions, or are recorded and then ignored. Inspectors will ask a child something, then look for the record of it, and see whether anything followed.

Missing-from-care records

Each episode, the return interview, and crucially the pattern. What did the home learn across six episodes, and what did it change? Six well-recorded episodes with no analysis is a weaker answer than three with a clear response.

Medication records

Checked for accuracy and for what happens after an error. A home that has recorded and reported its own medication errors is usually in a stronger position than one whose records show none at all.

Staff files, supervision and training

Recruitment checks, induction, supervision frequency, and whether training is current for the needs of the children actually placed. Supervision that happens but is not recorded counts, for these purposes, as supervision that did not happen.

Regulation 44 reports and the Regulation 45 review

Read as a set, and tested for whether recommendations were closed out. See Regulation 44 and Regulation 45.

Questions inspectors put to managers

These are not gotchas. They are all versions of the same question: does this manager know their own home?

• Tell me about the child you are most worried about at the moment, and what you are doing about it.
• What has got better here in the last six months, and how do you know?
• What is not good enough yet?
• What did your last Regulation 44 report say, and what happened as a result?
• How do you know your staff are recording well? What do you do when they are not?
• Which of these children would tell me they are listened to, and which would not?

The answer that goes badly is the one that describes systems. The answer that goes well names a child, names the concern, and names what happened next.

What they ask staff and children

Staff are usually asked what they would do if a child disclosed something, who they would go to, and how they know a particular child’s risks. Inconsistency between staff is itself the finding: it suggests the knowledge lives in individuals rather than in the home.

Children are asked, in their own words, whether they feel safe, who they would tell, whether anyone asks them what they think, and whether anything changed when they did. An inspector will then go looking for the record of that in the file.

What actually helps beforehand

Very little that can be done in the fortnight before an unannounced inspection makes much difference, which is the uncomfortable part. What does help is narrow:

1. Read two children’s files end to end as if you were the inspector, and fix what you find.
2. Close out every open Regulation 44 recommendation, or record honestly why one is still open.
3. Check supervision and training currency, because these are binary and easily evidenced.
4. Make sure the staff team can each answer the safeguarding question in their own words.

Common questions

Hearth is an independent software provider and is not affiliated with, endorsed by, or connected to Ofsted. This guide is general information about the regulatory framework, not regulatory or legal advice. Always check the current Children’s Homes (England) Regulations 2015, the Guide to the Children’s Homes Regulations and the SCCIF on GOV.UK for the authoritative detail.

← All guides