Restraint of trade clauses in Australian contracts: what they contain and how they work

You've just hired a senior salesperson or signed a contractor with full access to your systems and client list. Someone on the other side has sent back a draft employment agreement — or perhaps you're the one issuing it. Either way, your eye lands on the restraint clause. It's either a single, sweeping paragraph or a dense grid of cascading options. Neither version tells you what it actually does, whether it would hold up, or whether it's going after the right things.

Restraint of trade clauses are among the most commonly misunderstood provisions in commercial contracts. They're not a blanket prohibition on competition — courts won't enforce them that way. They're a targeted tool for protecting specific business interests after a relationship ends. Getting the drafting right requires understanding what each moving part does and where each one commonly fails.

What a restraint clause actually restrains

A restraint of trade clause restricts what a person can do after their engagement with your business ends. In an employment or contractor context, that typically means limiting post-exit conduct involving your customers, staff, confidential information, or competing activities. In a sale of business context, a restraint prevents the seller from immediately setting up shop next door and recapturing the goodwill you just paid for.

The legal starting point is that restraints are presumed void at common law — courts treat restrictions on freedom to work and trade as contrary to public policy unless they can be justified. A restraint will be enforced only if it is reasonably necessary to protect a legitimate business interest and goes no further than that interest requires. That principle applies across all Australian states and territories.

NSW adds a statutory layer: under s 4 of the Restraints of Trade Act 1976 (NSW), a restraint is valid to the extent it is not against public policy, and the NSW Supreme Court has power to read down an unreasonable restraint — treating it as enforceable to the extent that is reasonable — rather than striking it out entirely. Courts in other jurisdictions generally apply blue-pencil severance (cutting out void words) but cannot rewrite the clause.

The legitimate interest clause

Before any restraint can bite, the party enforcing it must identify a legitimate business interest that warrants protection. This is not a formality — courts scrutinise it.

Recognised legitimate interests include:

• Trade connections and goodwill — client relationships built up through your business, not merely through the individual's personal reputation
• Confidential information — pricing structures, product roadmaps, internal processes, supplier terms, and similar information that is genuinely confidential and not publicly available
• Workforce stability — preventing the systematic poaching of your team by a departing person who knows who to target

Skills and general industry knowledge are not a protectable interest. A restraint that is really just an attempt to stop someone from using their expertise will not be enforced. The clause should identify — either in its drafting or in the accompanying definitions — which specific interests it is protecting and why they are vulnerable given this person's role.

Trap: Omitting any express articulation of the interest. If a court cannot identify what the clause was designed to protect, it has nothing to calibrate the scope against. The restraint is more likely to be treated as an attempt to suppress competition generally, which will not survive scrutiny.

The activities clause

The activities clause defines what the restrained person is prohibited from doing. This is where most drafting goes wrong — either by being so broad it cannot be enforced, or so narrow it leaves gaps.

Common activity-based restraints include:

• Non-compete: prohibits the person from working in, owning, or advising a competing business. This is the hardest to enforce and requires the clearest justification.
• Non-solicitation of customers: prohibits the person from approaching your customers or prospective customers to take their business. More targeted and generally more enforceable than a full non-compete.
• Non-dealing: prohibits the person from doing business with your customers even if the customer approaches them first. More restrictive than non-solicitation and will be assessed accordingly.
• Non-poaching of staff: prohibits soliciting or inducing your employees or contractors to leave. Often the most commercially important clause for a scaling business.

For most small businesses, a non-solicitation and non-poaching regime — combined with strong confidentiality obligations — provides meaningful protection. A non-compete should be reserved for roles where the person genuinely could recreate your competitive position using insider knowledge, and even then, it must be scoped carefully.

Drafting minimum: Define each restricted activity specifically. "Competing business" needs a definition. "Customer" needs a definition (does it include leads? former customers? companies the person dealt with directly, or all your clients?). Vague activity clauses generate disputes about what was actually prohibited, which makes enforcement expensive and uncertain.

The time clause

The time clause sets how long the restrictions last after the engagement ends.

There is no universal correct duration. Courts assess whether the period is reasonably necessary given what is being protected:

• For customer non-solicitation, the relevant period is typically tied to your sales cycle, contract renewal cadence, and how long it would take to transition and stabilise the relationship with someone else
• For confidential information, the relevant period is how long the information will remain genuinely confidential and commercially useful — some technical or strategic information becomes stale quickly, while pricing structures or product roadmaps may remain sensitive longer
• For a non-compete, the period must be justified by reference to how long it would take a competitor to use the person's knowledge to damage your position

Trap: Defaulting to long periods because longer feels safer. Courts treat an excessive period as evidence that the clause is not a genuine attempt at proportionate protection — it reads as a punitive measure. A shorter, well-justified period is more likely to be enforced in full than a long one that invites a challenge.

The geographic scope clause

The geographic scope clause defines where the restraint applies. It should reflect where your business actually operates and where the person could realistically compete against you.

• A business with a defined local client base may justify a suburb- or city-level restraint
• A business that operates statewide may justify a state-level restraint
• For businesses that operate online nationally, a geographic restraint can be difficult to justify and may be better replaced with an activities-based restraint focused on your actual customer base

Trap: Applying a blanket Australia-wide restraint where the business has a defined service area. This pattern is often challenged successfully. If your customers are concentrated in metropolitan Sydney, an Australia-wide non-compete will draw scepticism. Consider using a specific geographic definition, or relying on a customer-defined non-solicitation instead of a location-based non-compete.

The definitions clause

A restraint clause depends on its definitions. Courts interpret restraints strictly, and an ambiguous definition is resolved against the party trying to enforce the clause.

The definitions that matter most:

• Customer: should specify the period during which the customer relationship was active, whether it includes prospective customers the person was involved in pursuing, and whether "customer" means the legal entity or individuals within it
• Confidential information: should be defined with specificity. Courts are reluctant to treat everything a person learned on the job as confidential. Focus on categories of genuinely sensitive information: pricing, supplier terms, technical specifications, unreleased product features
• Competing business: should describe the relevant industry or activity clearly enough to distinguish genuine competition from unrelated work that happens to be in a similar sector

Trap: Overly broad definitions. A definition of "confidential information" that sweeps in general industry knowledge will undermine the whole clause because it looks like an attempt to monopolise the person's skills rather than protect specific information.

The cascading (ladder) clause

A cascading or ladder clause sets out multiple versions of a restraint, each with different time periods or geographic scopes, structured from broadest to narrowest. The clause states that the first version applies to the extent it is valid, failing which the next applies, and so on.

This technique is widely used in NSW to work with the read-down power under the Restraints of Trade Act 1976 (NSW). Courts assessing a cascading clause can identify which version of the restraint is reasonable and enforce that version. Outside NSW, the availability of this approach depends on whether the clause is genuinely structured as separate, independent restraints (susceptible to blue-pencil severance) rather than a single clause with multiple options.

A cascading clause is not a substitute for thinking carefully about what is reasonable. Courts have treated poorly constructed cascading clauses as attempts to circumvent the reasonableness requirement, which can undermine the whole provision.

Optional and situational clauses

Depending on your business and the relationship being documented, consider whether the following additions are warranted:

• Garden leave provision: suspends the person on full pay during their notice period, keeping them out of the market without needing to rely on post-termination restraints, which carry more legal risk. Worth including for senior roles where the notice period itself has protective value.
• Acknowledgement of reasonableness: a clause in which the person acknowledges that the restraint is reasonable and that they received consideration for agreeing to it. Not a guarantee of enforceability, but can assist in establishing context.
• Survival clause: expressly states that the restraint obligations survive termination of the contract for any reason. Important where the contract might be terminated for cause and there is any risk the restraint could be read as falling away.
• IP and confidentiality tie-in: links the restraint clause expressly to the confidentiality and IP assignment provisions, reinforcing that the restraint exists to protect the specific interests addressed in those clauses.
• Jurisdiction clause: specifies the governing law, which matters in a restraint context because NSW courts have the read-down power under the NSW Act and other states do not.

How Artificer Legal approaches these clauses

Restraint clauses are regularly returned to us in a form that is either unenforceable as drafted or disproportionate to the risk being managed. The pattern we see most often: a template clause written for a large employer applied without modification to a small business, a junior hire, or a contractor relationship where it has no realistic chance of holding up.

When we review a restraint clause, we work through the interest being protected first — not the time period or geography. The scope questions only make sense once the legitimate interest is clearly identified. A clause built backwards from the time period tends to be a clause built to lose.

Where we are drafting the clause, we push for precise definitions before anything else, because that is where most disputes actually arise. We would rather spend time on a watertight definition of "customer" than on whether the restraint period is twelve or eighteen months.

Where we are reviewing a clause on behalf of someone being asked to sign, we assess the scope against their actual role and exposure. A clause that is unenforceable does not protect the employer — but a clause that is broader than necessary creates friction in negotiations and genuine uncertainty for the person signing it.

We also look at the broader contract context. A restraint sitting in an otherwise thin document — no clear confidentiality obligations, no IP assignment, no acknowledgement of consideration — is structurally weaker than one embedded in a properly drafted agreement that supports the restraint's purpose throughout.

The clause that determines whether the document works

The definitions clause decides more restraint disputes than any other part of the document. Courts interpret restraints strictly. Where a word is ambiguous, it is read against the enforcing party. The party challenging the restraint does not need to prove the clause is unreasonable across the board — they only need to find a plausible interpretation under which the clause sweeps more broadly than justified.

Time and geographic scope get the most attention during drafting, but it is the definition of "customer" or "competing business" that tends to be the pivot point in actual litigation. A restraint with well-calibrated time and geography, but a definition of "customer" that includes leads the person never met, will fail on that definition.

Restraint clauses are most effective when they are read not as a standalone provision but as the final layer of a coherent protection framework: IP ownership terms establish what belongs to the business, confidentiality obligations set the rules around sensitive information, and the restraint clause adds a targeted post-exit buffer for the period where exposure is highest. A clause that tries to carry all of that work by itself will almost certainly carry too much.