Skip to content
Close
Contact Us
Contact Us
Social Housing

Guide · Damp and mould

Find damp and mould risk before the report.

Most damp and mould is found when a resident reports it. By then the home has a problem and the Awaab's Law clock is running. This guide covers which data signals point to risk, how to join them at property level, how to use household data safely, and where to start.

Book a callSee the data signals

By Catalyst BI housing data teamLast reviewed 9th November 20269 minute read

6%of housing association homes had a damp problem in 2024English Housing Survey 2024 to 2025.
1 to 2%of social homes estimated to have serious damp and mouldRegulator of Social Housing, 2023.
27th Oct 2025damp and mould came under Awaab's Law timescalesWhere it presents a significant risk.

00The case for prediction

Why predict damp and mould instead of waiting for a report?

Because a report means a resident is already living with it. Since 27th October 2025, damp and mould that presents a significant risk has come under Awaab's Law timescales, which start when you become aware of a hazard. Prediction is about knowing sooner.

6%of housing association homes had a damp problem in 2024

The Regulator's best estimate in 2023 was that less than 0.2% of social homes had the most serious damp and mould, 1 to 2% had serious problems and a further 3 to 4% had notable problems (Regulator of Social Housing). The English Housing Survey found damp in 6% of housing association homes in 2024 (MHCLG).

The difference between landlords was data. Stronger responses to the Regulator showed robust data on the condition of tenants' homes. Weaker ones relied more on reactive approaches and had weaker data about the homes they manage. Some high performers went further, looking at similar properties once damp and mould was found in one home (Regulator of Social Housing).

Government guidance says the same: landlords "should also proactively collect information on the condition of their housing stock" (MHCLG, DHSC and UKHSA).

01Where the risk shows up

Which data signals point to damp and mould risk?

The causes are well known: condensation, leaks, poor ventilation and homes that are hard to heat. Most housing associations already hold signals for each one. They just sit in different systems.

7types of signal most associations already hold
SignalWhat it tells youUsually held inWatch out for
Archetype, age and constructionWhich homes are built in ways that trap moisture, such as solid walls or flat roofsAsset management systemArchetype fields left blank or set at stock transfer and never updated
Stock condition surveyThe state of windows, roofs, gutters, ventilation and heatingAsset management system, survey filesSurveys several years old; components recorded for some homes and not others
Repairs historyRepeat leaks, mould washes, condensation visits and extractor fan repairsRepairs system, contractor portalsVague job descriptions; work logged against a block, not a home
Contact and complaintsResidents mentioning damp, condensation or a smell before a formal reportContact centre, CRM, complaints systemFree text that needs classifying before it can be counted
Energy performanceHomes that are hard or costly to heat, where condensation builds upEPC records, asset systemExpired or estimated certificates
HouseholdOccupancy and circumstances that make a home harder to heat and ventilateHousing management systemSensitive data: see the section on using it safely
SensorsTemperature and humidity readings over timeSensor platforms, where installedOnly covers homes with sensors; needs the resident's agreement

Government guidance names condensation, leaks, inadequate ventilation and insufficient heating and insulation as common causes, and says smart sensors may be a useful aid (MHCLG, DHSC and UKHSA).

02Joining the data

How do you join the signals at property level?

Every signal has to land on the same home. That needs one property reference across your housing, asset, repairs and contact systems, and a clear rule for blocks, communal areas and shared roofs.

Start with the property. The Awaab's Law data standard developed by MHCLG with HACT is built around four entities: tenants, property, investigation appointments and work orders (MHCLG Digital). Modelling your data on it means the same joined-up records serve prediction and your Awaab's Law evidence trail.

Then deal with the text. Repairs descriptions and contact notes hold the earliest signals, but they are written by hand. AI can tag them, for example as condensation, leak or ventilation, for a person to check, so they can be counted alongside the structured data.

Finally, link homes to the buildings and parts of buildings they share. A leak from a shared roof is a risk to every flat under it, not just the one that reported it.

Questions to ask your team
  • Do our repairs, asset and housing systems use the same property reference?
  • Can we tell which flats share a roof, riser or communal heating system?
  • How many damp-related jobs are logged with a vague description?

03Scoring risk

What does a useful damp and mould risk score look like?

One a surveyor and a resident could understand. Start with clear rules built from the signals above, test them against past cases, then try a statistical model and keep it only if it does better.

  1. Define what you are predicting. For example, a confirmed damp and mould case in the next 12 months, using your own case records.
  2. Start with rules. A repeat condensation visit, no record of a working extractor fan and a high-risk archetype is a simple, explainable flag.
  3. Test it on the past. Run the rules on last year's data and check how many of this year's confirmed cases they would have flagged.
  4. Then try a model. A statistical model can weigh many signals at once. Keep it only if it finds more real cases than the rules, and you can still explain each result.
  5. Show the reasons. Every flagged home should list the signals behind its score, so the person reviewing it knows what to look at.
  6. Refresh it. Recalculate monthly as new repairs, contacts and surveys arrive, and check the results across patches and household types for unfair patterns.

A score is only as good as the data under it. Homes with missing survey data or vague repairs records will look lower risk than they are, so report data gaps alongside the scores.

04Using household data

How do you use household and vulnerability data safely?

Vulnerability should change how quickly and carefully you act, not whether a household gets help. Keep it out of the risk score where you can, and apply it afterwards, with a person making the decision.

Childrenand older adults are at greater risk of more severe health impacts

Government guidance says children and older adults are at greater risk of more severe health impacts from damp and mould, as are people of all ages with a weakened immune system (MHCLG, DHSC and UKHSA). That makes household information useful for deciding who to visit first. It also makes it sensitive.

  • Lawful basis and minimisation: use only the fields you need, for a purpose you have recorded, and say so in your privacy notice.
  • Health data: treat it as special category data, with tighter rules and tighter access.
  • Access: limit who can see vulnerability flags and log every view.
  • People decide: the score suggests, a trained person decides. Draft ICO guidance on automated decisions says human involvement must be active, not tokenistic (Slaughter and May).
  • Priority, not rights: use vulnerability to bring a visit forward or adjust how you communicate. Never use it to restrict a tenancy or a service.

Agree the approach with your data protection officer before you build. It is much easier to design in than to retrofit.

Questions to ask your team
  • Do we have a lawful basis for each household field we use?
  • Who can see vulnerability flags, and is that logged?
  • Does a person review every flag before anyone acts on it?

05Acting on it

What should happen when a home is flagged?

A flag is a prompt for a visit or a conversation, not a diagnosis. Route it into the processes you already run, and record what happens next on the same trail as a reported case.

  • Contact the resident with practical advice and an offer to inspect, in the way that suits the household.
  • Inspect the highest scores through your planned or proactive inspection programme.
  • Check similar homes in the same block or archetype once a problem is confirmed, as the Regulator found high performers doing.
  • Record it properly. If an inspection finds a hazard, log it the same way as a reported case, with the case reference and timestamps your evidence trail needs. Check with your legal adviser how the Awaab's Law timescales apply to hazards you find yourself.

Then measure whether it works: the share of new reported cases in homes you had already flagged, the time from flag to visit, and how often damp returns after works.

06First steps

Where should a housing association start?

With one patch or one archetype, the data you already hold and a three-month test. Prove the signals work for your homes before you scale.

  1. Pick the scope. One estate, one archetype or one area with a history of damp cases.
  2. List the signals you hold. Use the table above and mark each one as complete, partial or missing for that scope.
  3. Join them on one property reference. Fix the matches that fail before building anything else.
  4. Build the rules and back-test. See how many recent confirmed cases the rules would have flagged.
  5. Run it live for three months. Visit the top-scoring homes, record what you find and compare it with the reports that come in.
  6. Decide what next. Scale, add a model, or fix the data gaps the test exposed.

The same joined-up data supports your Awaab's Law evidence, TSM repairs measures and board reporting, so the work pays back more than once.

FAQs

Questions about damp and mould data.

Not covered here? Ask us directly.

Can you predict damp and mould in social housing?

You can rank homes by risk using data most housing associations already hold: archetype and construction, stock condition surveys, repairs history, contacts and energy performance. It will not catch every case, and it works best where stock condition and repairs data are complete and joined on one property reference.

What data do you need to predict damp and mould risk?

Start with property archetype and age, stock condition survey components, repairs history such as repeat leaks and condensation visits, contact centre and complaint mentions of damp, and EPC records. Household data and sensor readings can help, but need careful governance and the resident's agreement where relevant.

Does Awaab's Law require landlords to predict damp and mould?

No. Awaab's Law sets timescales for responding once a landlord is aware of a hazard. Government guidance and the Regulator do expect landlords to know the condition of their homes and to act proactively, and prediction is one practical way to do that.

Can we use vulnerability data in a damp and mould risk model?

Use it carefully. Keep it out of the risk score where you can and apply it afterwards to decide who to visit first, with a person making the decision. Treat health information as special category data, limit access and agree the approach with your data protection officer.

Do we need AI to predict damp and mould?

Not to start. Clear rules built from repairs, stock condition and archetype data are often the right first step, because they are easy to explain and test. A statistical model or AI that reads repairs notes can be added once the data is joined up, and kept only if it finds more real cases.

Sources

Every date and figure on this page comes from the sources below, checked on 9th November 2026.

  1. MHCLG, Awaab's Law: guidance for social landlordsTimescales for damp and mould and emergency hazards from 27th October 2025.
  2. Regulator of Social Housing, Initial findings on damp and mould in social housing2nd February 2023. Estimated share of homes with damp and mould, and reactive versus proactive approaches.
  3. Regulator of Social Housing, Damp and mould in social housing: learning the lessons28th June 2023. What high-performing landlords did with their data.
  4. MHCLG, DHSC and UKHSA, Understanding and addressing the health risks of damp and mould in the homeFirst published 7th September 2023, updated 1st April 2026. Causes, health risks and proactive data collection.
  5. MHCLG, English Housing Survey 2024 to 2025: headline findings on housing quality and energy efficiency29th January 2026. Damp by tenure in 2024.
  6. MHCLG Digital, Delivering a data standard to support Awaab's Law in social housing20th August 2025. The four entities of the standard.
  7. Slaughter and May, New ICO guidance on ADM: some relaxations but no free pass2026. Summary of draft ICO guidance on automated decision-making after the Data (Use and Access) Act 2025.

This guide explains the data side of damp and mould risk. It is not legal advice. Check how the rules apply to your organisation with your legal adviser and data protection officer.

Find your highest-risk homes first.

Tell us about one patch or archetype that worries you. We'll start by understanding your use case and the data you hold today, then come back with where the gaps are and a solution that fits how you work.

Book a callBack to Social Housing
Also for housing data teamsEvidence trail guide →STAIRs data guide →