SHARE Knowledge Hub

Health and Safety Policy

A health and safety policy explains an employer’s intent, named responsibilities and the arrangements that control real work. It must be written when there are five or more employees in Great Britain or Northern Ireland. Its value is tested at the workface: can people find, use and verify the arrangements?

Updated 2026-10-19UK focus

Understanding Health and Safety Policy

A health and safety policy explains an employer’s intent, named responsibilities and the arrangements that control real work. It must be written when there are five or more employees in Great Britain or Northern Ireland. Its value is tested at the workface: can people find, use and verify the arrangements?

Why this matters at work

Define what each check can and cannot prove. A completed inspection sheet may establish that someone visited a site, but not that every barrier worked. A low injury rate does not show that chronic exposure is controlled. Set frequency based on consequence and variability, and create a route for urgent bad news between formal reviews. Compare data sources when they conflict: a green dashboard beside repeated worker reports is a reason to investigate, not to discount the reports.

Management restricts the suspect work as needed, obtains competent exposure and ventilation advice, reviews worker reports and funds control repair or redesign. The board receives the uncertainty, interim protection, responsible owner and verification result. Surveillance or medical referral may be appropriate, but appointments do not replace exposure control. Revise indicator design to show critical extraction condition, exposure evidence and worker signals. Check whether the same blind spot exists for noise, vibration or stress.

How organisations manage this in practice

A final management review should integrate the risk profile, worker consultation, critical-control condition, incidents, occupational-health signals, contractor performance, audit findings, change and resource decisions. Record what the evidence says about the effectiveness of the policy, not only whether its pages are current. Preserve unresolved risk and uncertainty in the report, with interim controls and named decision-makers.

State who may accept a risk assessment, when work must wait for a control and how significant findings are communicated. The five-employee recording threshold for significant findings is related but distinct from the policy-writing requirement. A ticked assessment is weak evidence if it relies on unavailable equipment or an impossible sequence. Verify at the task that controls work, and review after change, incident, new knowledge or a reported defect. Record uncertainty where exposure or control effectiveness has not been established.

Secure the route, inform affected employers and workers, and test the traffic assessment against the new conditions. Identify who allowed storage in the walkway, how changes to delivery timing are approved and what inspections should have found the defect. Assign physical correction and verify it across shifts. Then update the relevant arrangement and policy wording if ownership or change controls were missing, communicate the change and sample other shared entrances. Record the reason for the review and the workface evidence that risk was reduced.

What is a health and safety policy?

A health and safety policy is a formal statement of intent, organisation and arrangements for managing health and safety risks. See Health and Safety Executive (HSE) guidance.

The useful way to understand a policy is as the architecture of the management system. The statement of intent says where the organisation is going. The responsibilities section says who owns what. The arrangements section explains the practical systems used to control risks, such as risk assessment, consultation, training, inspection, maintenance, incident reporting, emergency response, contractor control, audit and review.

A weak policy sounds positive but does not govern behaviour. It may say that the company is committed to safety, but it does not identify who signs off high-risk work, who manages contractors, how actions are closed, how workers are consulted or how senior leaders know controls are working. A strong policy makes decision rights and evidence visible.

For learners, the key is not to memorise the three headings. It is to explain how policy converts legal duty into a practical management system that can be implemented, monitored and improved.

Policy partPurposeGood content
Statement of intentSets leadership commitment and strategic direction.Signed commitment, aims, legal compliance, continual improvement and worker involvement.
Organisation/responsibilitiesAllocates who does what.Named roles, directors, managers, supervisors, workers, competent advisers and consultation routes.
ArrangementsExplains how risks are managed in practice.Risk assessment, controls, training, maintenance, emergency, inspection, monitoring, audit and review.
ObjectivesTurns intent into measurable improvement.SMART objectives linked to significant risks and assurance priorities.
Review processKeeps the policy current.Triggers, owner, version control, communication and management review.

How can the hierarchy of controls turn safety policy commitments into effective risk reduction?

Eliminate, Reduce, Isolate, Control, personal protective equipment (PPE) and Discipline (ERICPD) are prompts for choosing and checking controls. A policy is a management commitment, not a physical barrier. Consider a busy loading bay where pedestrians cross forklift routes: the source is moving vehicle energy and the pathway is the shared route. The policy must assign who selects, funds, installs, checks and revises the controls. HSE’s five levels are elimination, substitution, engineering, administrative controls and PPE; Eliminate, Reduce, Isolate, Control, personal protective equipment (PPE) and Discipline (ERICPD) adds the Reduce, Isolate and Discipline prompts as decision prompts. Neither makes a signed policy an adequate control by itself.

Eliminate: the site designer and operations lead test whether pedestrian access through the loading bay can be removed by relocating deliveries or creating a separate entrance. Verify that the alternative does not move people into a different vehicle route; record the design decision and owner.

Substitute and reduce: if crossing cannot be removed, change delivery timing or handling method to reduce vehicle movements and the number of people exposed. The manager verifies actual traffic counts and service impact. A shorter shift or a warning poster alone is not substitution.

Isolate and engineer: install a physical pedestrian route, barriers and controlled crossing points with adequate sightlines. Facilities and operations verify the barrier in place before opening the route and inspect it after vehicle strikes or layout changes. A missing barrier or blocked crossing stops the affected movement until safe separation is restored.

Control, PPE and Discipline: assign traffic rules, drivers’ and pedestrians’ instruction, supervision and defect reporting; high-visibility clothing may help visibility but cannot replace segregation. The policy states who maintains the route, records checks, acts on near misses and revises arrangements. An audit should observe the actual crossing, not only the signature page.

Direct source: HSE how to write your policy

Direct source: HSE policy law

Direct source: HSENI health and safety policy

Direct source: HSE hierarchy of controls

Changing delivery timing reduces exposure through administration. It is not itself substitution of the hazard; where no distinct safer replacement process is demonstrated, state that rather than forcing an example.

How does policy become an operating system?

A health and safety policy becomes useful when it creates a visible flow from intent to controls, evidence, assurance and improvement.

The strongest policies are not long for the sake of it. They create line of sight. A director should be able to trace a commitment in the policy through the risk register, into standards and procedures, into supervisor checks, into worker briefings, into action tracking and back into management review. If that line cannot be followed, the policy is probably decorative.

This is especially important for NEBOSH and Diploma learners. A good answer should show how policy, organisation and arrangements connect to HSG65's Plan, Do, Check, Act cycle. Policy sets direction, planning identifies risks and objectives, doing implements controls, checking tests evidence, and acting improves the system after learning.

Policy layerOperational translationEvidence that it is working
Statement of intentSenior leaders define priorities, legal compliance, worker involvement and continual improvement.Signed current policy, board minutes and resourcing decisions.
Risk profileSignificant risks are identified and ranked.Risk register, exposure data, incident history and consultation outputs.
ObjectivesPolicy aims become risk-linked SMART targets.Objective tracker with owners, dates and verification.
ArrangementsProcedures, standards, permits, inspections, competence and emergency arrangements are defined.Controlled documents, field checks and action closure.
ResponsibilitiesRoles know what they own, when to escalate and what evidence to keep.Responsibility matrix, appraisals and supervision records.
AssuranceLeaders test whether arrangements work in reality.Audit reports, critical-control checks and management review.
ImprovementIncidents, audits, worker feedback and change lead to updates.Revised arrangements, communication and version control.

What policy architecture keeps arrangements usable?

A mature policy architecture connects the top-level policy to standards, procedures, forms, records and assurance so people know what to do and leaders can prove it happened. See HSE how to write your policy.

One reason policies fail is that everything is pushed into one document. The policy should set direction and ownership, but detailed controls usually sit in supporting standards, procedures, permits, checklists, forms and records. This makes the system easier to maintain: when hot-work control changes, the hot-work standard and permit can be updated without rewriting the whole policy.

The important test is traceability. A board member should be able to ask how a policy commitment to control contractors appears in the contractor standard, prequalification form, induction record, site inspection, audit finding, action tracker and management review. If the chain breaks, the policy is not functioning as an operating system.

LayerPurposeExample
PolicySets intent, scope, legal commitment, leadership ownership and broad arrangements.Health and safety policy signed by the managing director.
StandardsDefine mandatory business rules for significant risk areas.Contractor management standard, lifting standard, COSHH standard or permit-to-work standard.
ProceduresExplain how a standard is applied in a task, site or process.How to approve a subcontractor or issue a hot-work permit.
Forms and toolsCapture consistent decisions and evidence.RAMS review form, inspection checklist, induction form or action tracker.
RecordsProve work was planned, controlled, checked and reviewed.Training matrix, inspection record, audit report, permit pack or maintenance log.
AssuranceTests whether policy and arrangements work in real conditions.First-line checks, second-line audits, third-line independent review and management review.

What is the difference between a weak policy and a strong policy?

A weak policy describes intentions; a strong policy governs risk, ownership and evidence.

A weak policy says “safety is everyone’s responsibility” without identifying who owns a hazardous route or how a failed control is repaired. A strong one names the decision owner, the work arrangement, the check and the route for worker feedback.

Policy featureWeak versionStrong version
IntentGeneric commitment to safety.Specific commitment to the organisation's significant risks, worker involvement and continual improvement.
ResponsibilitiesEveryone is responsible.Named roles, authority, escalation and evidence requirements are clear.
ArrangementsList of topics with no operating detail.Risk assessment, competence, monitoring, maintenance, contractor control and emergency systems are described.
ObjectivesEasy-to-count activity measures.SMART objectives linked to high-risk work, health risks and assurance findings.
ConsultationWorkers are informed after decisions.Workers and representatives influence risk controls and review.
EvidencePolicy exists on a noticeboard.Audits and observations show the policy is implemented in real work.
ReviewReviewed every few years by date only.Reviewed after change, incidents, audits, legal updates and risk-profile shifts.

What is the difference between a paper policy and a live policy?

A paper policy exists as a document; a live policy shapes decisions, resources, supervision, worker voice and assurance evidence.

A signed statement can be current on paper while a damaged loading-bay barrier remains in use. Test the policy by comparing its named responsibilities and inspection promise with the actual route, worker accounts and defect closure.

Test pointPaper policyLive policy
LeadershipSenior person signs once and rarely discusses it.Senior leaders use the policy to set priorities, ask assurance questions and resource controls.
ManagersManagers know the policy exists but cannot explain arrangements.Managers can explain their responsibilities, escalation rules and evidence requirements.
SupervisorsSupervisors rely on habit and local custom.Supervisors use policy-linked standards, pre-start checks and stop-work rules.
WorkersWorkers are told the rules after decisions are made.Workers help shape arrangements through consultation, feedback and reporting.
ContractorsContractors receive the policy but not practical expectations.Contractor standards, induction, permits and supervision make expectations visible.
EvidenceThe policy is available on a noticeboard or intranet.Audits, checks, incidents and action trackers show the arrangements are applied.
ReviewReviewed for a signature/date refresh.Changed when risk profile, work methods, incidents, law or assurance findings change.

How should policy objectives and SMART targets be written?

Policy objectives should be specific, measurable, achievable, relevant, time-bound and linked to significant risk.

A common weakness is choosing objectives because they are easy to count rather than because they reduce serious harm. 'Complete 100 inspections' may look active but can be meaningless if inspections do not test critical controls. A better objective might be to verify all high-risk lifting plans, reduce overdue statutory inspections, close all critical COSHH actions, or train supervisors in permit handback and isolation verification.

Objectives should be owned, resourced and reviewed. They should also be balanced. A policy that rewards only low injury numbers can suppress reporting. A mature policy includes leading indicators, lagging indicators, assurance quality, action closure and worker feedback.

Objective typeWeak wordingStronger wording
Risk assessmentReview risk assessments.Review and re-brief the 20 highest-risk task assessments by 30 September, with supervisor verification.
TrainingDo more training.Assess competence for all forklift operators and close gaps by site by 31 October.
InspectionComplete monthly inspections.Verify critical controls for work at height, lifting and COSHH using ranked findings and action closure.
ConsultationImprove communication.Run quarterly worker forums and track actions from concerns to close-out.
Health riskImprove wellbeing.Reduce uncontrolled wood dust findings through extraction checks, monitoring and refresher training.
AssuranceAudit better.Complete second-line audits of contractor control at all high-risk sites and verify corrective actions.

Illustrative completed objective: baseline 6 uncontrolled wood-dust findings in 30 sampled cutting tasks in September. By 31 January 2027, the workshop manager will demonstrate effective extraction/suppression in every sampled task, recording failures per tasks sampled across shifts and arranging an immediate hold/correction on any failed control. A second-line reviewer checks a fresh sample and closure evidence by the deadline. Reporting concerns remains encouraged; fewer reports is not the target.

What should leaders ask about policy assurance?

Leaders should test whether policy commitments are live in the organisation, not simply whether the document has been signed.

Leaders should ask whether critical controls exist and work, whether workers can report failures, and whether resources and corrective actions match the risk. A count of policy acknowledgements alone cannot answer those questions.

Assurance questionWhy it mattersEvidence to request
What are our top health and safety risks and do policy arrangements control them?Policy should be risk-profile led.Risk register, assurance map and critical-control verification.
Who owns each critical arrangement?Unclear ownership creates drift.Responsibility matrix and named action owners.
What evidence shows supervisors apply the policy?Field implementation is where policy succeeds or fails.Site observations, inspection records and worker interviews.
What are workers telling us that the policy does not yet address?Consultation should influence the system.Safety committee minutes, worker feedback and action close-out.
Which actions are overdue and why?Overdue actions show resource or ownership weakness.Action tracker, escalation record and management decisions.
What has changed since the policy was last reviewed?Change can invalidate arrangements.MOC, organisational change, new sites, plant, contractors or legal updates.

How should responsibilities be described?

Responsibilities should describe decision rights, accountabilities and evidence, not just job titles.

A responsibilities section should help people answer: what am I accountable for, what decisions can I make, when must I escalate, what evidence must I keep, and how will performance be reviewed? This matters because many incidents happen in gaps between departments, contractors, shifts or project phases.

For example, a director may own resources and governance, a line manager may own implementation, a supervisor may verify controls before work starts, a worker may use equipment correctly and report defects, and a competent adviser may support but not own operational risk. The policy should make that distinction clear.

RoleCore responsibilityEvidence
Board/directorsSet direction, resources, governance and risk appetite.Signed policy, board minutes and assurance questions.
Senior managersTranslate policy into business plans and standards.Objectives, budgets, risk reviews and action escalation.
Line managersImplement arrangements and manage competence, contractors and controls.Team plans, inspections and performance reviews.
SupervisorsBrief, verify and stop work where controls are not in place.Toolbox talks, pre-start checks and intervention records.
WorkersFollow training, use controls and report concerns.Engagement, defect reports and consultation feedback.
Competent advisersAdvise, monitor, challenge and support improvement.Reports, audits and technical guidance.

How can a RACI table clarify policy responsibilities?

A RACI table helps avoid blurred ownership by showing who is responsible, accountable, consulted and informed for key arrangements.

RACI is useful where policy responsibilities cross functions. For example, contractor control may involve procurement, project managers, site supervisors, HSE advisers, directors and the contractor's own management. Without a responsibility model, everyone assumes someone else has checked competence, RAMS, induction, permits or close-out.

The table should not replace judgement. It should make accountability visible enough that people can challenge gaps before work starts. For high-risk arrangements, the accountable person should have authority, resources and access to reliable assurance evidence.

ActivityResponsibleAccountableConsultedInformedEvidence
Set contractor standardHSE manager and procurement leadOperations directorProject managers, worker reps and legal/commercial supportAffected managers, supervisors and contractor contacts before the revised standard takes effect.Controlled standard and board approval where required.
Prequalify contractorProcurement and project managerContract ownerHSE adviser and technical authoritySite lead and contractor contact before appointment or mobilisation.Approved contractor record and competence evidence.
Review RAMS and interfacesProject manager and site supervisorPrincipal contractor/site leadContractor supervisor, HSE adviser and affected tradesAffected trades, workers and contractor supervisor before high-risk work.RAMS review, interface plan and pre-start briefing.
Supervise high-risk workSite supervisorProject managerContractor supervisor and workersWorkers, contractor contact and operations on changes or stop decisions.Inspection records, permits and stop-work actions.
Verify close-outContract ownerOperations/project managerHSE adviser and maintenance/facilities ownerOperations, affected users and contractor contact before handback or restart.Handback record, defects closed and lessons learned.

The responsible function sends each update at the stated decision point and checks receipt before release. RACI means Responsible, Accountable, Consulted and Informed. These allocations clarify work; they cannot transfer the statutory duties of employer, client, contractor or principal contractor.

What should the arrangements section cover?

Arrangements should explain the practical systems used to manage risk in the organisation.

The arrangements section is where the policy becomes operational. It should not be a list of good intentions. It should tell people how the organisation plans, organises, controls, monitors and reviews health and safety. This includes the systems that convert risk assessments into field controls and evidence.

A high-risk business needs arrangements for topics such as contractor control, permit-to-work, isolation, COSHH, maintenance, lifting, work at height, occupational health, emergency response and statutory inspection. A low-risk office still needs arrangements for DSE, fire, first aid, stress, facilities and incident reporting.

ArrangementWhat to explainEvidence expected
Risk assessmentHow risks are identified, assessed, reviewed and communicated.Risk register, task assessments and review records.
CompetenceHow competence is defined, assessed, maintained and verified.Training matrix, authorisations and observed competence.
ConsultationHow workers influence decisions and raise concerns.Forum minutes, safety rep actions and feedback loop.
ContractorsHow contractors are selected, briefed, supervised and assured.Pre-qualification, induction, permits and performance reviews.
MonitoringHow inspections, audits, KPIs and assurance are planned.Inspection/audit schedule and action closure.
IncidentsHow events are reported, investigated and used for learning.Reports, root cause analysis and lessons learned.
ReviewHow policy and arrangements are updated after change.Management review and version control.

What does a useful policy look like in practice?

A useful policy creates a line of sight from top-level commitment to work control and assurance.

Consider a contractor that has expanded from ten workers to 150 across multiple construction sites. The old one-page policy says everyone should work safely, but it does not explain who appoints subcontractors, who reviews RAMS, how supervisors verify controls, how plant defects are escalated or how directors know high-risk work is controlled. As the business grows, informal knowledge is no longer enough.

The updated policy should set out senior ownership, project-level responsibilities, arrangements for CDM, contractor control, competence, lifting, work at height, temporary works, occupational health, monitoring and review. It should also define the evidence that returns upward, such as audit results, critical actions, high-potential near misses and overdue inspections.

Old approachRisk createdImproved policy control
Everyone responsible for safety.No decision ownership.Define director, manager, supervisor, worker and adviser accountabilities.
Generic commitment to comply.No link to significant risks.Set objectives for high-risk work, health risks and statutory inspections.
No contractor section.Contractors unmanaged.Add contractor selection, induction, supervision and assurance arrangements.
No review trigger.Policy becomes stale after growth or change.Add review after incidents, organisational change and legal updates.
No evidence standard.Paper policy cannot be audited.Define records, KPIs, inspections, audits and management review.

Operational example and hold points

Task material and setting: Write and apply a warehouse loading-bay policy where vehicles and pedestrians may share a route.

People and prerequisites: Senior policy owner, warehouse manager, drivers, pedestrians and worker representatives; route plan and traffic observations available.

Before: Identify the vehicle–pedestrian conflict, consult users and test a separated route and named inspection responsibility.

During: Supervisors verify barriers and crossing controls; workers report a damaged or blocked route.

Hold or stop: Missing barrier, unsafe crossing, unowned inspection or a policy arrangement that cannot be used in practice.

Completion and hand back: Restore and verify segregation before reopening the affected movement; communicate any changed arrangement.

Evidence: Signed policy, named responsibilities, route assessment, field check, near-miss review, corrective action and revision record.

Where do health and safety policies commonly fail?

Policies fail when they are disconnected from risk, ownership, field evidence and review.

Policy failure often begins when arrangements are copied from another business, responsibility is unassigned or change is not communicated. Check the actual work and revise the policy route when new tasks, equipment, contractors or incidents reveal a gap.

FailureHow it shows upWhat to improve
Template wordingPolicy says the right things but does not match the business.Rewrite around real risk profile and operations.
No ownershipActions fall between departments.Name roles and escalation points.
Objectives detached from riskFocus on easy counts, not serious controls.Link SMART objectives to risk register and critical controls.
No consultationWorkers do not recognise the arrangements.Add worker involvement and feedback loop.
Not communicatedSupervisors and contractors cannot explain policy expectations.Use induction, briefings and manager training.
No reviewPolicy ignores new sites, plant, processes or law.Define review triggers and version control.
No assuranceLeaders cannot tell if arrangements work.Use audits, verification and management review.

What evidence shows a policy is effective?

Evidence should show that policy commitments are implemented, checked and improved.

Evidence should be read as a chain rather than a filing exercise. A single record rarely proves good management on its own; stronger assurance comes from matching records, observations, interviews, trends and verified action closure.

EvidenceWhat good looks likeWeakness to challenge
Signed policySigned by most senior person and reviewed regularly.Outdated signature or unknown owner.
CommunicationWorkers and contractors know relevant arrangements.Policy only on noticeboard.
ObjectivesSMART, owned and risk-linked.Objectives not tracked.
ResponsibilitiesPeople can explain their role and escalation routes.Generic 'all staff' wording.
ArrangementsPractical systems exist for main risks.Arrangements do not match work as done.
MonitoringInspections, audits and KPIs test policy effectiveness.Monitoring counts activity only.
ReviewPolicy updated after change, incident and audit learning.Policy unchanged for years.

When should the policy be reviewed or revised?

Policy review should happen after meaningful change, evidence of weakness or new information, not only because a calendar date has arrived.

Review after a material process change, new legal requirement, repeated control failure, serious event or evidence that workers cannot use an arrangement. Record the decision, owner, version and communication so the revised policy reaches the workface.

TriggerWhy it mattersPolicy review question
New work activity, site or processExisting arrangements may not cover the risk.Do responsibilities, competence, emergency and monitoring arrangements still fit?
Organisational restructureOwnership and escalation routes may change.Have named roles, authority and consultation routes been updated?
Serious incident or high-potential near missThe policy may not be driving effective control.Which arrangement failed and what must change?
Audit or enforcement findingFormal evidence shows the system is weak.Does the policy need clearer ownership, standards or assurance?
Worker feedback or repeated complaintsWork as done may differ from policy as written.Are arrangements practical and understood?
Legal or official guidance updateDuties or expectations may have changed.What needs updating and how will workers be told?
Growth, acquisition or contractor model changeInformal arrangements may no longer be adequate.Does the policy cover new interfaces and governance?

How should the control be managed and reviewed?

For a GB workplace, HSE guidance supports identifying the actual risk, taking practical protective steps and reviewing whether they work. Name the decision owner, check the control during the real task and correct a failed arrangement. Apply the separate Northern Ireland duties and guidance where the work is in NI.

GB management guidance: HSE managing health and safety

What training and awareness helps?

Training should help people understand their role in the policy, not just where the policy is stored.

Training should be role-based. General awareness creates a baseline, but supervisors, managers, advisers and senior leaders need different levels of decision-making, verification and escalation competence.

RoleDemonstrationAssessorRecord and reassessment
Workers and representativesFind the policy arrangements for their task, use consultation route and report a failed control.Supervisor checks a live task example.Induction and change communication; record feedback.
SupervisorsTranslate policy into pre-use checks, instruction, stop decisions and closed corrective action.Manager observes the control on the worksite.Task check and reassessment after change or incident.
Policy owners and senior leadersApprove intent, assign named responsibilities and resources, test whether arrangements work in practice.Board or independent assurance lead tests evidence.Signed statement, review trigger and action closure.
Advisers and auditorsCompare policy words to risk assessment, worker interviews, field controls and monitoring.Competent reviewer samples actual operations.Assurance report and revision recommendation.

What do the acronyms and key terms mean?

Acronyms should be understood before they are used in assessments, briefings, audits or learning answers.

Acronyms can exclude newer readers if they are not explained. This section keeps the main terms available at the end of the page so the article stays readable while still supporting learners and audit users.

AcronymMeaningPlain-English use
HSWAHealth and Safety at Work etc. Act 1974Core Great Britain health and safety legislation; NI uses the separate 1978 Order.
SMARTSpecific, Measurable, Achievable, Relevant, Time-boundObjective-setting approach.
ISO 45001Occupational health and safety management systems standardInternational OH&S management-system framework.