The AI Journal

In the spring of 2014 the Singapore Court of Appeal considered a joint-venture agreement that required KS Energy Services to use “all reasonable endeavours” to procure the construction of an oil rig within six months. When the deadline passed and the rig remained unbuilt, the dispute turned on a question that has occupied commercial lawyers for decades: does “all reasonable endeavours” demand less than “best endeavours”? The court’s answer was blunt. Any differences between the two standards, it held, were “more metaphysical than practical”, and the test for compliance should ordinarily be the same for both. That judgment ought to have unsettled the drafting profession. If the adjectives make no practical difference, why do negotiators continue to treat them as distinct rungs on a ladder of obligation?

The conventional wisdom holds that common law jurisdictions recognise a three-tier hierarchy. “Best endeavours” sits at the top, sometimes said to require a party to sacrifice its own commercial interests; “all reasonable endeavours” occupies a middle ground, requiring the exhaustion of all reasonable avenues without necessarily demanding such sacrifice; and “reasonable endeavours” is the baseline, satisfied once a single reasonable course of action has been pursued. This taxonomy is so deeply embedded in practice that standard-form contracts routinely deploy all three formulations in different clauses, on the assumption that the words themselves calibrate the intensity of the obligation.

Yet the case law of the past decade has eroded the supposed gap between the top two tiers. In KS Energy Services Ltd v BR Energy (M) Sdn Bhd, the Singapore Court of Appeal held that both “best endeavours” and “all reasonable endeavours” clauses require the obligor to take “all those reasonable steps which a prudent and determined man, acting in the interests of the obligee and anxious to procure the contractually-stipulated outcome within the available time, would have taken”. The same formulation had earlier been advanced in England, but the Singapore court went further in explicitly collapsing the distinction between the two phrases. Benjamin Wong notes that it is now settled in Singapore that the two clauses “generally impose the same standards on the obligor”. The English position is not markedly different. In Rhodia International Holdings Ltd v Huntsman International LLC, the High Court observed that an obligation to use all reasonable endeavours “is likely to equate to using best endeavours”, while in Jet2.com Ltd v Blackpool Airport Ltd the Court of Appeal treated the two formulations as interchangeable for the purposes of the dispute before it. The result is that the top two rungs of the traditional hierarchy have merged into a single standard of exhaustive but reasonable effort.

What remains genuinely distinct, most authorities agree, is the gap between that merged standard and “reasonable endeavours” simpliciter. A “reasonable endeavours” clause is generally understood to be discharged by pursuing just one reasonable course of action, whereas the higher standard requires the exhaustion of all reasonable avenues. But even this distinction is less stable than it appears. Context repeatedly overrides the label. In Electricity Generation Corp v Woodside Energy Ltd, the High Court of Australia held that sellers who had undertaken to use “reasonable endeavours” to supply extra gas were entitled to refuse when market prices spiked, because reasonableness permitted them to weigh their own commercial interests. Conversely, in Jet2.com the English Court of Appeal held that a “best endeavours” clause could require an airport to operate at a financial loss where the contract’s central purpose depended on it. The decisive factor in each case was not the adjective preceding “endeavours” but the contractual objective and the commercial matrix. As Greatrex observes, the meaning of these clauses is “fluid and context-specific”, and pre-formulated distinctions between “best”, “all reasonable” and “reasonable” distort the obligation by inviting courts to split semantic hairs rather than assess what was reasonable in the circumstances.

The persistence of three separate formulations therefore generates costs without corresponding benefits. Parties spend negotiating time haggling over adjectives that the courts will later treat as synonymous or near-synonymous. When disputes arise, the court’s first task is to determine which of the three boxes the clause belongs to, but its second task—applying the standard—inevitably returns to the same question: what would a reasonable person have done? The adjective adds little to that inquiry. Indeed, the Singapore Court of Appeal in KS Energy Services described the effort to distinguish “best” from “all reasonable” as a “pointless hair-splitting exercise”. If that is right, then the continued use of both phrases in the same contract is not sophisticated drafting but contractual noise.

The strongest objection to this view is that parties choose words deliberately, and that a textualist approach to interpretation requires the court to give effect to the parties’ choice of adjective. If a contract uses “best endeavours” in clause 6 and “all reasonable endeavours” in clause 9, the objection runs, the drafters must have intended a gradation of obligation; to collapse the two is to rewrite the contract. This argument has force where the same instrument employs both formulations for different obligations. Yet it assumes that the drafters had a shared, legally cognisable understanding of the difference—and the evidence suggests they do not. Empirical studies of contract practice are lacking, but the judicial record is replete with cases in which sophisticated parties have argued for distinctions that the courts have struggled to articulate. The problem is not merely that the difference is hard to pin down; it is that the difference does not reliably produce different outcomes. A court asked to enforce a “best endeavours” clause still asks what steps were reasonable; a court asked to enforce an “all reasonable endeavours” clause asks the same question. The adjective does not supply a metric. It merely invites litigation.

A more defensible position would reserve “reasonable endeavours” for the lower standard—one reasonable course—and treat “best endeavours” and “all reasonable endeavours” as a single, higher standard requiring the exhaustion of all reasonable measures. That is, in substance, what the courts are already doing. The Singapore guidelines endorsed in KS Energy Services and elaborated in subsequent case law apply identically to both formulations: the obligor must continue until all reasonable endeavours are exhausted, need only pursue steps with a real prospect of success, and is not invariably required to sacrifice its own commercial interests unless the contract so contemplates. These guidelines do the real work; the adjective does not. As Tan Tian Yi observed, the guidelines “strike a suitable balance between the interests of both contracting parties” because they shift attention away from the label and towards the concrete steps a reasonable obligor should take.

If the distinction between “best” and “all reasonable” endeavours has become metaphysical, the practical response is to recognise the merger explicitly. Drafters who wish to calibrate obligation should do so by specifying the steps required—appeal the planning decision up to the Court of Appeal, spend no more than £50,000, maintain the airport’s normal hours unless losses exceed £10,000 per month—rather than by varying an adjective that the courts have shown they will ignore. The recent United Kingdom Supreme Court decision in MUR Shipping BV v RTI Ltd [2024] UKSC 18, discussed by Christopher Hose, reinforces the same principle: a reasonable endeavours proviso is concerned with securing contractual performance according to its terms, not with substituting different performance. The law of endeavours clauses would be clearer, and contracts would be more enforceable, if the profession abandoned the pretence that “best” and “all reasonable” describe different things. They do not. The sooner drafting practice acknowledges what the courts already know, the less money will be wasted on arguments that begin and end with the dictionary.