Random drug and alcohol testing in Canadian workplaces: The evolving legal landscape

Introduction

Employers have a fundamental obligation to ensure that every worker is provided with a safe workplace. While employers are not expected to police employees’ off-duty conduct, that interest may shift when off-duty conduct impacts workplace safety. The law recognizes this tension and continues to evolve to address the realities of modern workplaces.

For employers operating in safety-sensitive industries, the legal framework presents both opportunities and constraints. Understanding where the boundaries lie is essential for developing compliant and defensible workplace substance-use policies.

Four key decisions define the current state of the law:

1. Communications, Energy and Paperworkers Union of Canada, Local 30 v. Irving Pulp & Paper, Ltd., 2013 SCC 34 (Irving) – the preeminent Supreme Court of Canada decision establishing the balancing test for random testing in unionized workplaces.
2. Power Workers’ Union et al. v. Attorney General of Canada et al., 2024 FCA 182 (OPG) -the Federal Court of Appeal decision upholding random drug testing in a nuclear safety context, which potentially expands the Irving framework.
3. Amalgamated Transit Union, Local 113 v Toronto Transit Commission, 2026 CanLII 72133 (TTC) – an Ontario labour arbitration decision pulling the analysis squarely back to Irving and its core requirement of demonstrated workplace evidence, even in a public transit safety context.
4. Dome Mine Limited and United Steelworkers, Local 7980, 2026 CanLII 78264 (Dome Mine) – an Ontario labour arbitration decision upholding post-incident testing in the mining sector and clarifying the limited scope of the pre-testing investigation required under Irving.

This article examines each of these decisions in turn and provides practical guidance for employers considering random testing programs.

The Irving standard: The foundational balancing test

Communications, Energy and Paperworkers Union of Canada, Local 30 v. Irving Pulp & Paper, Ltd., 2013 SCC 34 (Irving) is the seminal, foundational case at the Supreme Court of Canada to establish the balancing act employers must take when conducting random testing in a unionized workplace.

In Irving, the Supreme Court of Canada addressed whether a paper mill could unilaterally impose a mandatory random alcohol testing policy for employees in safety-sensitive positions. Under the challenged policy, 10% of employees in safety-sensitive positions were to be randomly selected for unannounced breathalyzer testing over the course of a year. In the 15 years that preceded the introduction of this policy, there were only eight documented incidents of alcohol consumption or impairment at the mill. In that time, there were no accidents, injuries or near misses connected to alcohol.

The balancing test

Following a grievance filed by an employee, the arbitrator weighed the employer’s legitimate safety interest in random alcohol testing against the harm to employees’ privacy interests. Ultimately, a majority of the Supreme Court agreed with this balancing approach, holding that random alcohol testing constitutes a significant intrusion on employee privacy and that an employer must justify that intrusion by demonstrating a sufficient safety need.

Critically, the Court held that a dangerous or safety-sensitive workplace does not, on its own, justify random testing. The employer must demonstrate something more to support random testing; typically, evidence of a demonstrated workplace problem with substance use that has not been mitigated through other means.

The KVP framework

The scope of management’s unilateral rule-making authority under a collective agreement is set out in Re Lumber & Sawmill Workers’ Union, Local 2537, and KVP Co. (1965), 16 L.A.C. 73. Under the “KVP test,” any rule or policy unilaterally imposed by an employer and not subsequently agreed to by the union must be consistent with the collective agreement and be reasonable. Random testing must satisfy these requirements, and the “reasonableness” criterion is where the balancing test applies.

What Irving permits

Irving only permits substance testing in narrow, targeted circumstances without requiring evidence of a broad workplace problem.

These include:

• Reasonable-cause testing (where there are objective indications of impairment).
• Post-incident testing (following a significant workplace incident).
• Return-to-work and monitoring testing (as a condition of continued employment following a substance-use admission or incident).
• Pre-employment testing (in certain limited circumstances).

Only random testing requires the heightened justification of demonstrated workplace need.

The FCA’s approach in OPG: Random testing in the nuclear context

In Power Workers’ Union et al. v. Attorney General of Canada et al., the Federal Court of Appeal (FCA) addressed random drug testing in the uniquely regulated nuclear power sector; specifically, Ontario Power Generation (OPG). The case concerned the legality of a random oral fluid testing program for employees in nuclear safety-sensitive positions.

The nuclear regulatory context

The FCA’s analysis was based not on a unilaterally instituted employer policy, but instead upon the distinctive statutory framework of the Nuclear Safety and Control Act. The purpose of that legislation is to limit risks posted by nuclear energy to national security, the health and safety of persons and the environment. The regulatory regime imposes stringent obligations on licensees to ensure fitness for duty among personnel performing safety-critical functions. This context provided a statutory foundation quite different from that in Irving’s paper mill.

The FCA’s holding

The FCA upheld the random drug testing regime. Significantly, the Supreme Court of Canada subsequently dismissed the application for leave to appeal with costs, leaving the FCA judgment intact as the final word on this matter.

The FCA found that the uniquely safety-critical, highly regulated nuclear workplace justified enhanced safety requirements that tipped the Irving balance in the employer’s favour. In the nuclear context, the catastrophic consequences of any safety failure would be far-reaching, with potential harm to the public, the environment and national security. These elements elevated the employer’s safety interest to a degree that outweighed employee privacy concerns, even without the same workplace-specific substance abuse evidence that Irving required.

Key distinction from Irving

The critical distinction lies in the nature of the regulatory environment and the magnitude of potential harm. Irving involved a conventional industrial workplace that, while dangerous, limited that danger largely to the site itself. OPG involved a facility where a single lapse could result in radiological harm extending far beyond the workplace itself. The FCA’s reasoning suggests that the more catastrophic the potential consequences, and the more comprehensive the statutory and regulatory framework demanding fitness for duty, the stronger the employer’s justification for preventive measures including random testing.

Post-OPG uncertainty

Following OPG, the decision appeared to signal a broader willingness to accept random testing in safety-sensitive environments without evidence of a demonstrated workplace substance-use problem. Whether OPG represented a narrow, context-specific application of the Irving balance, confined to the nuclear regulatory context, or a more expansive reading remained uncertain. Recent decisions from arbitrators are now answering that question.

Dome mine: Post-incident testing and the limited investigation requirement

In Dome Mine Limited and United Steelworkers, Local 7980, Arbitrator Bendel addressed the termination of a heavy equipment operator who refused to submit to post-incident drug and alcohol testing following a serious safety violation at a gold mine in Timmins, Ontario.

The facts

The grievor and a co-worker were assigned to work at the mine’s Coarse Ore Stockpile area, which required a “full lockout” by the mill team before work could begin. Both employeesfailed to comply with the lockout procedure, although they mistakenly believed at the time that they had done so. They worked for approximately three hours before another employee noticed the violation and told them to leave the area immediately. The employer’s investigation report concluded that the lockout violation “resulted in workers being exposed to uncontrolled hazards and energy sources” and “potential significant injury from active feeders conveying coarse ore.”

The employer ordered post-incident testing under its Fit for Work: Alcohol and Drug-Free Workplace Program. The grievor refused, stating there was no “probable cause” for such testing. His co-worker submitted to testing and tested positive for impairment. The grievor was terminated for refusing the testing order.

The scope of pre-testing investigation

The union argued that the employer had failed to conduct an adequate investigation before ordering the testing. Arbitrator Bendel rejected this argument, endorsing the Ontario Labour Relations Board’s analysis in Mammoet Canada Eastern Ltd. that the pre-testing investigation need only establish “a link between the incident and the employee’s situation.” Crucially, the arbitrator held that management is not required to have a reasonable basis to suspect actual impairment before ordering testing:

“[T]he line of inquiry must focus on whether there is a link between the incident and the employee’s situation. If the supervisor concludes that the acts or omissions of the employee could not have been a contributing factor, then testing is not required… It is not reasonable to expect a supervisor, while conducting an investigation in a short window of time, to accurately assess whether an employee is impaired by drugs or alcohol.”

This is a significant clarification. Post-incident testing under Irving does not require a “reasonable cause” assessment of impairment; rather, it only requires that the employee was involved in a significant incident. The purpose of the testing is to rule out impairment as a contributing factor, not to confirm suspected impairment.

Refusal to test as a terminable offence

Arbitrator Bendel upheld the termination, accepting the employer’s argument that reinstating an employee who refused testing would “incentivize” future refusals. Employees who know or suspect they are impaired might calculate that they have nothing to lose by refusing the test and might keep their jobs if an arbitrator substitutes a lesser penalty based on mitigating factors such as long service. For this reason, the arbitrator endorsed treating a refusal to submit to legitimate post-incident testing as equivalent in seriousness to a positive test result.

Consistency and past practice

The union also argued that the employer had not ordered testing following other, similar safety violations in the past. Arbitrator Bendel held this was “wholly irrelevant” to whether post-incident testing was justified in the present case. The employer’s statutory obligation to take “every precaution reasonable in the circumstances for the protection of a worker” under the Occupational Health and Safety Act cannot be defeated by pointing to past failures to take such precautions. However, the arbitrator noted that past inconsistency could be relevant to the quantum of discipline.

Trachuk and the TTC: Pulling back to Irving

In TTC, Arbitrator Trachuk addressed random drug testing in the public transit context through a case involving the Toronto Transit Commission (TTC). While transit operations are undeniably safety-sensitive, the question was whether that characterization alone could justify random testing of employees.

The overall message

Arbitrator Trachuk reaffirmed the Irving framework and its central requirement that there must be clear evidence of a drug and alcohol problem in the workplace before random testing can be justified. Despite the inherent safety-sensitivity of TTC operations, with hundreds of thousands of passengers transported daily on buses, streetcars and on the subway, the arbitrator held that the dangerousness of the work environment alone is insufficient to justify the privacy intrusion of random testing.

Distinguishing OPG

The TTC decision effectively limits OPG to its unique facts and the specific regulatory requirements underpinning the nuclear sector. Arbitrator Trachuk rejected any broader reading of OPG that would allow random testing in all safety-sensitive environments merely because the work involves risk to the public. Transit operations, while undeniably important and safety-sensitive, do not carry the same catastrophic, radiological and national-security risks that made the nuclear context exceptional, nor are transit operations subject to specific statutory or regulatory requirements that might otherwise justify a departure from the Irving principles.

The evidence requirement

The decision underscores that the Irving evidentiary threshold remains alive and well for the vast majority of Canadian workplaces. An employer seeking to implement random testing must come to the table with concrete, persuasive evidence of a substance-use problem actually manifesting in its workplace. General statistics about substance use in society, anecdotal concerns, or the theoretical possibility of impairment are insufficient. The evidence must demonstrate that the specific workplace has experienced a problem that less invasive measures have been unable to address.

Practical significance

For employers in safety-sensitive industries outside the nuclear sector, TTC is a clear signal that the path to lawful random testing runs through either: (i) the evidence-based Irving framework; or, (ii) legislative reform. If an employer as large and safety-sensitive as the TTC cannot justify random testing without demonstrated evidence of a workplace problem, employers in less safety-critical industries face an even steeper burden.

Key takeaways for employers

Irving, OPG, TTC and Dome Mine provide the following practical guidance for employers considering drug and alcohol testing programs:

1. The Irving balancing test remains the governing framework. For the overwhelming majority of Canadian workplaces, random testing must be justified by evidence of a demonstrated substance-use problem in the specific workplace. Safety-sensitivity alone is not enough.
2. OPG is narrow, not a green light. The FCA’s decision in OPG should be understood as a context-specific application of the Irving balance in an exceptional regulatory environment. It does not stand for the proposition that all safety-sensitive employers can implement random testing without workplace-specific evidence.
3. Build your evidentiary record. Employers who believe random testing may be necessary should diligently document workplace incidents, near-misses, positive test results from other forms of testing, referrals to employee assistance programs and other indicators of substance-use issues. This evidence will be essential to defending a random testing program.
4. Targeted testing remains available. Reasonable-cause testing, post-incident testing and return-to-work/monitoring testing remain lawful and appropriate tools. Employers should ensure these programs are well-designed and consistently applied. Evidence from targeted testing programs may also contribute to the evidentiary basis for random testing if a problem is demonstrated.
5. Consider the regulatory context. Employers subject to a specific regulatory framework mandating fitness for duty (such as nuclear, aviation or rail safety regimes) may have a stronger basis for random testing, particularly if the regulatory scheme itself contemplates or requires such measures. However, this is a fact-specific analysis.
6. Less invasive alternatives matter. Arbitrators will ask whether less privacy-invasive means could address the safety concern. Robust education programs, supervisor training in recognizing impairment, clear policies with meaningful consequences and employee assistance programs should be in place before resorting to random testing.
7. Unionized employers face additional constraints. Random testing implemented through unilateral policy (rather than negotiated into the collective agreement) must satisfy the KVP framework.
8. Refusal to test may be treated as seriously as a positive result. Arbitrators have recognized that allowing employees to escape consequences by refusing testing would “incentivize” refusals, particularly by those who suspect they would fail. Employers should ensure their policies clearly state that refusal to submit to a lawful testing order will be treated as equivalent to a positive test result.
9. Post-incident testing requires only a link to the incident, not suspected impairment. As Dome Mine confirms, the pre-testing investigation for post-incident testing need only establish that the employee was involved in the incident. Management is not required to suspect actual impairment before ordering testing. The purpose is to rule out impairment as a contributing factor, not to confirm it.
10. Past inconsistency does not bar current compliance. The fact that an employer failed to order testing following previous incidents does not preclude it from ordering testing now. The statutory duty to protect workers cannot be defeated by pointing to past failures. However, inconsistent enforcement may be relevant to the severity of discipline.

Conclusion

Canadian law on random workplace drug and alcohol testing remains firmly anchored in the Irving balancing test. While OPG demonstrated that exceptional contexts exist where the balance may tip in favour of random testing even without workplace-specific evidence of a substance problem, TTC has reaffirmed that for the vast majority of employers, including those in safety-sensitive industries, the evidentiary requirement remains paramount. At the same time, Dome Mine confirms that post-incident testing, one of the targeted testing mechanisms that Irving endorsed, remains available and does not require suspected impairment before it can be ordered. Employers should approach random testing cautiously, invest in building their evidentiary foundation and ensure that their broader substance-use policies and testing programs (reasonable cause, post-incident, return-to-work) are comprehensive and defensible. Any employer contemplating the implementation of a random testing program should seek specific legal advice in light of this evolving jurisprudence.