Drug, Alcohol and Fitness for Work: The Policy Every Prequalification Asks For and Few Contractors Can Defend

Somewhere in almost every civil prequalification application and most tender returnables, there is a request for your drug and alcohol policy. Most contractors satisfy it the same way: find a template, put the company name on it, sign the front page and upload it. It passes. It usually keeps passing for years.

The document only matters twice. The first is when a client audits it, or asks a question in an interview that the template cannot answer. The second — the one that actually costs money — is the morning a supervisor forms a view that someone should not be operating an excavator, and nobody in the business knows what the policy says to do next.

A fitness for work policy is an operational procedure that happens to be requested as a document. This guide covers what it needs to contain, the testing decisions that sit behind it, and the two directions the legal risk runs — because a contractor who tests badly is exposed at least as much as a contractor who does not test at all.

Why it is asked, and what is being tested

Clients ask for three reasons, and knowing which one is operating tells you how much detail your answer needs.

• Their own duty. A principal has work health and safety duties toward workers on their site, including yours. They discharge part of that duty by requiring contractors to manage impairment, and by being able to show they required it.
• Site access control. On many sites — rail corridors, mines, ports, airside areas, some major projects — a compliant policy plus testing is a condition of entry, not a scoring criterion. Without it your people cannot get through the gate.
• A capability signal. For everyone else, the policy is read as evidence of whether you run a managed business. An evaluator cannot audit your culture, so they look at whether your documents suggest one.

The third case is the most common in ordinary council and state civil work, and it is where a generic template does the least damage and the least good. It is also where a well-constructed policy is a cheap differentiator — most of your competitors submitted the same template you nearly submitted.

The duty underneath the paperwork

Australian work health and safety law places a primary duty of care on a person conducting a business or undertaking to ensure, so far as is reasonably practicable, the health and safety of workers. Workers carry a corresponding duty to take reasonable care for their own safety and not to adversely affect others.

Nothing in that framework specifically requires drug and alcohol testing. What it requires is that you manage the risk of impairment, and testing is one control among several. That distinction matters more than it sounds, because it means two things at once: a contractor who does no testing at all is not automatically in breach, and a contractor who tests everyone regularly has not automatically discharged the duty. The question is always whether the risk is being managed so far as is reasonably practicable, given the work.

For civil construction the risk profile is not marginal. The work involves mobile plant, excavation, working near live traffic, lifting, and work near services and overhead lines — the hazards covered in our guides to temporary works and excavation support and working near overhead powerlines. Impairment in those environments has consequences that are immediate and not recoverable. That is the reason the topic is treated seriously by clients, and the reason a policy that exists only as a PDF is a genuine exposure rather than an administrative gap.

The regulator in your jurisdiction publishes guidance on managing impairment at work, and several publish material specific to construction. Work from that guidance rather than from a template of unknown origin, and where your work is covered by a separate regime — rail safety work, mining, aviation, heavy vehicle driving — the requirements of that regime override general guidance and are usually far more prescriptive.

Fitness for work is broader than drugs and alcohol

A policy titled “drug and alcohol” covers a subset of the actual risk. Clients increasingly ask for fitness for work, which is a wider concept and a better framing for your own purposes.

Factor	How it presents on a civil site	Typical control
Alcohol	Residual from the night before more often than consumption at work	Policy, testing, education about clearance times
Illicit drugs	Rarely visible; detected rather than observed	Policy, testing, support pathway
Prescription and over-the-counter medication	The most common genuine impairment issue, and the one policies handle worst	Disclosure obligation, task modification
Fatigue	Long shifts, night work, travel to remote sites, second jobs	Rostering, journey management, hours limits
Heat and dehydration	Summer earthworks, sealing, confined and enclosed work	Work scheduling, hydration, acclimatisation
Acute medical and mental health	Undisclosed conditions, personal crisis, bereavement	Supervisor awareness, referral, task modification

Writing the policy as a fitness for work policy does two useful things. It gives supervisors a framework for the situations they actually encounter — of which “I think he has been drinking” is far from the most frequent — and it gives you a better answer to a client than a testing regime alone. An evaluator who reads a document that anticipates the medication problem and the fatigue problem is reading something written by a business that has thought about it.

What a defensible policy actually contains

Ten components. A policy missing any of the first six will not survive a serious audit, and more importantly will not tell a supervisor what to do.

• Scope. Who it applies to — employees, labour hire, subcontractors, visitors, management — and where, including travel to site and company vehicles.
• The standard. What is expected: presenting fit for work, not consuming during work hours, not being under the influence, and the specific position on the site.
• Obligations. What a worker must do, including disclosing medication that may affect capacity, and what a supervisor must do when they have a concern.
• Testing. Which situations trigger a test, who conducts it, to what standard, and how a worker is selected where testing is random.
• The stand-down procedure. What happens between a concern being raised and a result being known — including how the person gets home, because putting someone you believe to be impaired into a vehicle is a decision the policy must have already made.
• Consequences. What follows a confirmed positive result, and how that differs between a first occurrence, a refusal, and a repeat.
• Support. The pathway for someone who self-discloses a problem, and whether self-disclosure is treated differently from detection. A policy with no support pathway guarantees that nobody will ever disclose anything.
• Confidentiality and records. Who sees results, where they are kept, how long for, and how privacy obligations are met.
• Education. How workers are informed — at induction, in toolbox talks, and when the policy changes.
• Review. When the policy is reviewed, and by whom.

One additional element separates a real policy from a downloaded one. Name the roles. Not “management will arrange testing” but the position that decides, the position that arranges, and the position that is called when the decider is unavailable at 6am on a Saturday. Policies fail at the point where nobody is sure whether they have authority to act.

The four testing situations and what each one proves

Testing regimes are built from four situations. Each has a different purpose, a different level of intrusion, and a different evidentiary weight.

Situation	Purpose	What to watch
Pre-employment	Screening at the point of engagement	Sets a baseline expectation; must be applied consistently to avoid discrimination exposure
Random	Deterrence — the main value is that it might happen, not what it finds	Selection must be genuinely random and documented; a “random” test that always lands on the same person is not random and will be characterised as targeting
For cause / reasonable suspicion	Response to observed behaviour or performance	The highest-risk category. Requires trained supervisors and a contemporaneous written record of the observations that triggered it
Post-incident	Part of investigating what happened	Trigger threshold must be defined in advance — every incident, or only those meeting stated criteria — and applied without discretion

Two practical notes. Post-incident testing needs to be settled before an incident, not during one, and it must sit alongside the other immediate obligations covered in our guide to incident notification, site preservation and investigation. For a notifiable incident, the site preservation obligation and the testing decision run at the same time, and it is much easier if the answer is already written down.

And reasonable suspicion testing is where policies most often fail, because it depends entirely on supervisor capability. A supervisor who cannot describe what they observed, in specific and behavioural terms, has created a decision that will not survive challenge. Training supervisors to observe and record — speech, coordination, behaviour, performance, appearance — rather than to conclude, is the practical control here.

Detection is not impairment — the gap that causes disputes

This is the single most important technical fact about drug testing, and most policies are silent on it.

A drug test detects the presence of a substance or its metabolites. It does not measure whether a person is impaired. For alcohol the two correlate closely enough for practical purposes. For most other substances they do not. Detection windows vary by substance, by test type, by dose, by frequency of use and by individual physiology, and for some substances a positive result is possible well after any effect has passed.

The consequences for how you write and apply a policy are direct.

• Say what your policy is actually about. If the standard is “no detectable substance”, say so — that is a legitimate position on a high-risk site, but it is a different rule from “not impaired” and should not be presented as the same thing.
• Do not describe a positive result as proof of impairment in your documents, your investigation reports or your correspondence. It is not, and characterising it that way weakens your position if the matter is ever tested.
• Expect the argument. The gap between detection and impairment is the most common basis on which a testing decision is challenged, particularly in relation to cannabis, where detection windows can be long.
• Choose the test type deliberately, because different sample types have materially different detection windows — the point of the next section.

Testing methods and the standards behind them

Australian Standards cover the collection and analysis of specimens for drug testing — separate standards apply to urine and to oral fluid — and a further standard covers breath alcohol testing devices. Your policy should name the standards you test to, and your testing provider should be able to demonstrate compliance with them. This is the detail that distinguishes a considered policy from a template, and it is a fair question to ask a provider before engaging them.

Sample type	General characteristics	Practical implication
Breath (alcohol)	Immediate, non-invasive, closely related to current effect	The simplest to administer on site and the least contested
Oral fluid	Shorter detection window; collection is observed and less intrusive	Often preferred where the intent is to detect recent use; increasingly the construction default
Urine	Longer detection window; collection raises privacy and dignity issues	Detects use further back in time, which widens the detection-versus-impairment gap

Three procedural points that matter more than the choice of sample type.

• An on-site screening device gives a preliminary result only. A non-negative screen must be sent to an accredited laboratory for confirmatory analysis, and no consequence should attach until that confirmation is returned. Acting on a screening result alone is the most common serious procedural error.
• Chain of custody is the whole evidentiary basis. Sample identification, sealing, custody records and transport must be intact. A break in the chain makes the result unusable, whatever it says.
• Use a competent provider. Collection by an untrained person, however well intentioned, undermines everything downstream.

This guide does not state cut-off concentrations, detection windows in hours or days, or thresholds for any substance. Those figures are set in the relevant standards and by the testing regime that applies to your work, they change, and applying a number from a general article to a real decision about a real person is exactly the error this section is warning against. Take the numbers from the current standard and from your testing provider, not from an article.

Prescription and over-the-counter medication

This is the most frequent genuine fitness for work issue on a civil site and the one most policies handle badly, usually by ignoring it or by demanding disclosure of all medication — which creates a privacy and discrimination problem while producing information nobody is qualified to interpret.

A workable approach has four parts.

• Frame the obligation around effect, not diagnosis. The worker’s duty is to disclose that they are taking something that may affect their ability to perform safety-critical work — not what it is or what it is for.
• Put the question to the prescriber. The practical mechanism is a form describing the person’s actual tasks — operating plant, working at height, driving, working near live traffic — which the treating practitioner or pharmacist assesses. You receive a fitness opinion, not a medical record.
• Have somewhere for the person to go. Temporary task modification is usually straightforward on a civil site. If the only available outcome is being sent home unpaid, workers will not disclose, and the policy has produced the opposite of its intent.
• Keep it confidential and separate. Health information attracts privacy obligations. It should not sit in a general personnel file accessible to supervisors.

There is a related exposure worth naming. A worker with a disclosed medical condition or an addiction may have protections under anti-discrimination law and, if managed badly, a claim. The presence of a health condition changes what a fair process looks like, and this is a point at which employment advice is genuinely warranted rather than a formality.

What happens after a non-negative result

The sequence should be written before it is needed, because every step is one where a business without a procedure improvises badly.

• Remove the person from safety-critical work immediately. This is a safety step, not a disciplinary one, and should be described as such.
• Arrange transport. Do not let them drive. The policy should say who arranges it and who pays.
• Send the sample for confirmation and wait. No consequence, no announcement, no conversation with the client until the result is confirmed.
• On confirmation, put the result to the worker and give them a genuine opportunity to respond, including the opportunity to explain a legitimate cause such as prescribed medication.
• Then decide, consistently with how you have decided in comparable cases, and record the reasoning.

Two decisions to make in advance rather than in the moment. Is a first confirmed positive automatically dismissal, or is there a return-to-work pathway? Both are defensible positions; what is not defensible is deciding case by case in a way that varies with how much you value the individual. And how do you treat a refusal to test? Most policies treat refusal as equivalent to a positive result, which is reasonable, but it only works if the policy says so before the refusal happens.

Procedural fairness and the dismissal risk

The risk in this area runs in two directions, and contractors routinely see only one of them. The first is the safety risk of an impaired worker. The second is the employment law risk of a dismissal that is found to be unfair — and in this area, dismissals are challenged and are often successful, not because the substance was not detected, but because of how the process was run.

The failures cluster in a short list.

• No policy, or a policy never communicated. A rule the worker was never told about is difficult to enforce.
• Inconsistent application. A history of tolerating conduct in one person and dismissing another for the same conduct is the most damaging fact pattern available to the other side.
• Acting on a screening result without laboratory confirmation.
• No opportunity to respond before the decision was made.
• Procedural defects in collection — untrained collector, broken chain of custody, no second sample.
• Disproportionate outcome relative to the role, the risk, the person’s record and the policy’s own stated consequences.

Consistency is the control that does most of the work here, and it is free. It also has a bearing on other parts of your business: an unfair dismissal finding is a matter you may be asked to disclose, and a pattern of employment disputes eventually shows up in the same places as a poor safety record — including in the prequalification questions covered in our guide to civil contractor prequalification in Australia. Before dismissing anyone on the basis of a test result, take employment advice. It is a small cost against the alternative.

When the client’s policy is not your policy

On most sites you will encounter a client policy that sits over yours, and the interaction produces predictable friction.

• The client’s standard may be stricter than yours. Their site, their rule. Your policy should state that where a site standard is more stringent, the site standard applies.
• The client may test your workers. Establish before mobilisation who conducts it, who receives the result, and what the client may do with it. A client removing your worker from site is a decision about site access; whether it is also a decision about that person’s employment is yours, and confusing the two creates an unfair dismissal risk you have inherited.
• Your subcontractors are your exposure. A subcontractor’s worker on your site is your safety problem and, in the client’s view, your contractual problem. Flow the requirement down, verify it rather than accepting an assurance, and cover it in your subcontract terms.
• Costs. Client-mandated testing regimes have real cost — testing fees, lost time, replacement labour when someone is stood down. On a site with a heavy regime, this is a preliminaries item, not an overhead absorption.

Resolve these at mobilisation rather than at the first incident. Our guide to contract award and mobilisation covers the wider set of interface questions that should be settled before work starts.

Fatigue: the fitness issue with a separate legal spine

Fatigue belongs in a fitness for work policy, but for civil contractors it also sits inside a separate and more prescriptive regime. If your business operates heavy vehicles, drivers’ work and rest hours are regulated under heavy vehicle law, and the obligation extends beyond the driver to every party in the chain — including you as the party scheduling the work. Setting a delivery time that cannot be met without breaching work and rest limits is itself a breach, regardless of who was behind the wheel. That regime is covered in our guide to chain of responsibility for civil contractors.

Beyond the regulated driving hours, the general WHS duty applies to fatigue in the ordinary conduct of civil work, and the risk factors are structural rather than personal.

• Night and shutdown work. Road, rail and utility work is frequently done at night because the asset cannot be closed in the day. The programme is set by the client’s constraint, and fatigue management has to be built into it — see writing a construction programme for tenders.
• Travel time to site. Long unpaid drives at each end of a shift extend the effective day well beyond the hours worked, and are almost never counted.
• Remote and camp-based work. Compressed rosters concentrate hours by design — see remote community infrastructure and mining and resources civil works.
• Recovery work after weather or an event. The pressure to catch up is exactly when hours extend and controls lapse.

Implementing it in a business that has never had one

Introducing testing into a workforce that has never had it is a change management exercise, and doing it badly produces a policy that is resented, evaded and eventually unenforceable.

• Consult before you implement. WHS law requires consultation with workers on matters affecting their health and safety, and this plainly qualifies. It is also the step that determines whether the policy is accepted.
• Give notice and educate first. Announce the policy, explain it, explain the medication provisions and the support pathway, and let a stated period pass before testing begins. An amnesty period is common and sensible.
• Start with the leadership. A policy that visibly applies to directors and managers as well as operators is accepted; one that applies only to people in high-visibility clothing is not.
• Train supervisors properly, particularly on observing and recording rather than diagnosing.
• Document that you did all of this. The consultation record, the education record and the acknowledgement forms are the evidence that the policy was communicated — which is the first thing anyone will ask for.

Then keep it alive. A policy reviewed once and never mentioned again drifts into the same category as the template it replaced. Put it in induction, revisit it in toolbox talks, and review it when your work profile changes — a contractor moving into rail, mining or airside work is moving into regimes with their own mandatory requirements.

What tenders and prequalification actually ask

The questions are narrower than the subject, and they repeat.

What is asked	What a weak answer looks like	What a strong answer looks like
Attach your drug and alcohol policy	An undated template with another company’s terminology in it	A current, dated, signed policy with your roles named in it
Describe your testing regime	“We conduct random testing”	The four situations, the sample type, the standards, the provider, and how random selection is made
How do you manage fitness for work?	A restatement of the drug and alcohol policy	The wider set — medication, fatigue, heat, acute conditions — with the controls for each
How do you ensure subcontractor compliance?	“It is a condition of our subcontract”	The condition, plus how it is verified before mobilisation and monitored during the work
Evidence of implementation	Nothing, or the policy again	Induction records, toolbox records, consultation record, testing summary statistics

That last row is where most contractors lose available marks. The policy is the least interesting thing you can submit, because every bidder has one. What separates responses is evidence that it operates — and that evidence already exists in your induction and toolbox records if anyone thinks to attach it. The general principle in our guide to addressing selection criteria applies directly: the document is the claim, the record is the proof.

Checklist

• Is the policy current, dated, signed and written for your business rather than a template?
• Does it cover subcontractors, labour hire and visitors as well as employees?
• Are the decision-making roles named, including who is called outside business hours?
• Does it distinguish detection from impairment, and state which standard you are applying?
• Are the four testing situations defined, with the trigger for post-incident testing fixed in advance?
• Is random selection genuinely random, and is the method documented?
• Are supervisors trained to observe and record rather than to conclude?
• Does the policy name the Australian Standards your testing is conducted to?
• Is confirmatory laboratory analysis required before any consequence attaches?
• Is there a written stand-down procedure, including how the person gets home and who pays?
• Does the medication provision ask about effect on tasks rather than diagnosis?
• Is there a support and self-disclosure pathway that a worker would actually use?
• Are health records kept separately and confidentially?
• Have you decided in advance whether a first confirmed positive is dismissal or a return-to-work pathway?
• Is refusal to test addressed in the policy?
• Was the policy consulted on, communicated and acknowledged — and can you produce that record?
• Does fatigue management address night work, travel time and remote rosters, and align with your heavy vehicle obligations?
• Do you know what the client’s site policy requires, who tests, and who receives the result?
• Have client-mandated testing costs been priced as a preliminaries item?
• Can you attach implementation evidence to a tender, not just the policy?

Sources and further reading

This guide is general information for Australian civil construction businesses and is not legal, employment, medical or work health and safety advice. It deliberately states no cut-off concentrations, detection windows or testing thresholds for any substance; those figures are set by the applicable Australian Standards and by the regime governing your work, they change, and they must be taken from the current standard and your testing provider rather than from an article. Work health and safety duties, privacy obligations, anti-discrimination law and unfair dismissal law differ between jurisdictions and apply to particular facts. Decisions about a worker’s employment following a test result carry real legal risk and should be made with employment law advice. Consult your work health and safety regulator’s current guidance on impairment at work, and where your work falls under a separate regime — rail safety, mining, aviation or heavy vehicle law — comply with that regime’s specific requirements.

• The model work health and safety framework as enacted across Australian jurisdictions — the primary duty of care of a person conducting a business or undertaking, the duties of workers and officers, and the consultation requirement referenced in §02 and §13. Impairment is managed under the general duty rather than by a specific testing obligation, which is the distinction drawn in §02. Each jurisdiction’s regulator publishes current guidance on managing impairment at work; that guidance, not this article, is the operative source.
• Australian Standards govern the collection and analysis of specimens for drug testing, with separate standards for urine and for oral fluid, and a further standard covers breath alcohol testing devices. These are referenced in §07 by role rather than by content; cut-off levels, detection windows and procedural detail sit in the current editions of those standards and are deliberately not reproduced here.
• Heavy vehicle work and rest hour requirements and the extension of duties to all parties in the chain, referenced in §12, are sourced in full in our guide to chain of responsibility for civil contractors. Sector-specific regimes — rail safety work, mining, and aviation-related work — impose their own mandatory testing requirements that override general guidance; the relevant sector guides are rail civil works, mining and resources civil works and airports, ports and freight civil works.
• The procedural fairness observations in §10 describe the categories of failure that commonly arise in employment disputes over testing decisions, presented as general reasoning rather than as a summary of decided cases. Nothing in that section is a substitute for advice on a particular dismissal.
• Related TenderBuilt guides carrying the primary-source detail referenced above: WHS management plans and SWMS, incident notification and investigation, workers compensation and injury management, temporary works and excavation support, working near overhead powerlines, civil contractor prequalification, contract award and mobilisation, enterprise agreements and labour rates and addressing selection criteria.