• Lawyers in Sydney
• Lawyers in Melbourne
• Lawyers in Brisbane
• Lawyers in Perth
• Family lawyers
• Employment lawyers
• Criminal lawyers
• Wills and estates lawyers
• Find a lawyer

The earlier appeal allowed the Builder’s appeal in part. The Court decided that the Owners had to give credit for part of the settlement amount, addressing the concern about double recovery: M Johnson Building Pty Ltd v O’Toole [2026] ACTCA 8 at [107], as summarised in this judgment at [1]. The present decision did not reconsider the underlying building dispute. It addressed the appeal costs and the Owners’ request for the appellate court to fix costs from the first-instance proceeding and quantify the related credit.

At first instance, the Builder had been ordered to pay the Owners’ costs on an indemnity basis, in two parts: two thirds of those costs up to the settlement with the Certifier on 20 April 2023, and all of them after that date ([2]). The parties agreed that this first-instance costs order should not be disturbed ([3]). The Court therefore treated the outstanding question as the costs of the appeal ([4]).

The parties sought different costs outcomes. The Builder argued that the Owners should pay 50 per cent of its appeal costs, relying on its success on one of the grounds argued and the Owners’ refusal to concede that ground ([5]). The Owners argued that the Builder should pay two thirds of their appeal costs. They pointed to three matters: the Builder abandoned a challenge concerning wet areas shortly before the appeal hearing, failed on the excavation-damages issue, and succeeded only on the settlement-credit issue ([6]). They also argued that the successful point arose only after the primary judgment and relied on evidence not properly before the first-instance court ([7]). Both sides alternatively submitted that each should bear its own appeal costs ([8]).

The Owners additionally sought orders fixing first-instance costs. They wanted to make the settlement credit more definite and avoid further dispute about its amount ([9], [14]). The Court declined to decide that question in the appeal, while leaving open other ways it might be resolved ([16]–[17]).

Why the court decided this way

The Court started with the general costs principles set out in Wu v Wu (No 3) [2024] ACTCA 35 at [8]. As described in this judgment, those principles include that costs orders are compensatory and that the discretion to award costs is to be exercised with fairness in mind ([10]). The judgment does not set out a comprehensive test for every case involving partial success. Its reasoning is tied to the outcomes and practical features of this appeal.

The Court assessed success across the appeal as a whole. The Builder had succeeded on the legal issue of double recovery, and that success materially contributed to the appeal being allowed in part. But the Court considered the excavation issue to be the dominant issue, and the Owners had succeeded on it ([11]). It also treated the costs of the wet-areas issue, which the Builder abandoned, as significant enough not to ignore ([11]). The decision therefore does not treat success on any single ground as automatically determining the costs outcome.

The Court then considered whether costs could be divided issue by issue. It accepted that the issues were independent, but concluded they were not discrete in a way that would allow the Court to isolate particular periods of work as referable to a particular issue ([12]). That distinction mattered. The fact that grounds were separate did not, on this record, make it workable to carve out identifiable blocks of appeal costs.

Against that background, the Court rejected both parties’ primary positions and adopted their shared alternative position. It reasoned that, where each side had partial success in different ways, it was not appropriate both to deprive each of part of its own costs and to require each to pay a portion of the other’s costs ([12]). It concluded that fairness supported an order that each party bear its own costs ([13]). The formal orders confirmed the first-instance costs order and made that appeal-costs order ([18]).

The Court dealt separately with the request to fix the first-instance costs and quantify the credit. It acknowledged the practical merit in resolving the issue and avoiding further costs ([16]). The Owners had provided a detailed breakdown and supporting material, on the assumption that the existing indemnity costs order would remain in place ([14]–[15]). However, the Court considered that the dispute about the process and the amount involved questions it should not resolve in this appeal. Determining those questions at appellate level could deprive the parties of the ability to challenge factual findings about the amount payable, because the only avenue of appeal would be by special leave to the High Court ([16]).

The Court therefore declined to fix the costs as part of the appeal. It identified alternatives that remained available: an application in the first-instance proceeding, assessment in the usual way, or further correspondence that might enable agreement ([17]). Those are possibilities recorded by the Court, not findings that any one route must be used.

Before and after state of the law

The judgment describes the general costs principles it applied by reference to Wu v Wu (No 3) [2024] ACTCA 35 at [8]. In particular, it identifies the compensatory nature of costs orders and fairness as relevant to the exercise of the costs discretion ([10]). It does not state that it is changing or replacing those principles.

The decision’s contribution is its application of those principles to a particular pattern of mixed success. The Builder succeeded on an issue that materially affected the earlier appeal result, but lost on the dominant excavation issue. The costs associated with an abandoned issue also mattered. The Court considered the overall proceeding rather than awarding costs mechanically by counting successful grounds ([11]–[13]).

The judgment also illustrates a limit on issue-based apportionment. Even though the appeal issues were independent, the Court found that they were not sufficiently discrete to permit time-based carving out of costs ([12]). That conclusion depended on the way the appeal and its issues were presented, as described in the judgment. It should not be read as a universal rule that costs can never be separated by issue.

For first-instance costs, the judgment emphasises the procedural implications of resolving disputed amounts at appellate level. The Court considered that factual findings about costs and the corresponding credit could be difficult to challenge if determined in this appeal, given the special-leave limitation it identified ([16]). Its decision was to leave the question open to processes at first instance or possible agreement, rather than to determine the amount itself ([17]).

The judgment thus records an application of existing costs principles, not a stated alteration to the law. It does not specify the exact amount of the credit, determine a new method for calculating it, or set out a general formula for allocating costs in every appeal involving partial success.

Key passages with plain-English translation

At [10], the Court said that the general principles in Wu v Wu (No 3) include “the compensatory nature of costs orders” and a discretion exercised “with fairness in mind”. In plain English, the costs question was approached as one of fair compensation and discretionary judgment, rather than as an automatic penalty based solely on who won a particular point.

At [11], the Court observed that “[t]he parties each had a degree of substantive success on the appeal”. It then identified the Builder’s success on double recovery and the Owners’ success on the dominant excavation issue. In plain English, neither side had won the appeal overall in a way that resolved the costs question by itself.

The Court also said at [11] that the costs of the abandoned issue were “significant enough that they should not be ignored”. This indicates that abandoning a ground did not make the work and expense associated with it irrelevant to the overall costs assessment.

At [12], the Court stated that the issues were independent, but “none were discrete in a way that would permit carving out” periods of costs attributable to particular issues. In plain English, the Court could distinguish the legal issues but could not reliably divide the appeal work into separate cost blocks.

At [13], the Court concluded that “fairness dictates” that the parties’ alternative positions be adopted, meaning each would pay its own appeal costs. This was the result of the combined assessment of partial success, the dominant issue, the abandoned issue and the difficulty of apportioning costs.

Finally, at [16]–[17], the Court explained why it would not fix the first-instance costs in the appeal. Although finalising the issue might avoid further expense, the Court considered that deciding disputed costs and related factual matters at appellate level could restrict the parties’ ability to challenge factual findings. It left the question to a possible first-instance application, assessment, or agreement.

What fact patterns trigger this precedent

The decision is relevant to costs questions where both sides achieve meaningful but different forms of success in an appeal. Its facts include success by the appellant on one material legal issue, success by the respondents on the dominant issue, and an abandoned issue whose costs remained significant ([11]).

It is particularly informative where a party asks for costs to be divided by issue or by the time spent on each issue. The Court’s reasoning turns on whether the issues are not merely independent, but sufficiently discrete to permit a reliable allocation of costs ([12]). If the costs cannot be separated in that way, this judgment shows that an overall assessment may be more appropriate on the facts.

The decision also concerns a situation where a party asks an appellate court to fix costs incurred at first instance and determine an associated credit, but the process or amount is disputed and may require factual findings. The Court’s stated concern was that making those findings in the appeal could limit the parties’ ability to challenge them ([16]). It identified possible first-instance procedures and assessment as alternatives, without prescribing which one should be used in another case ([17]).

The judgment does not establish that every mixed-success appeal should result in each party bearing its own costs, or that an appellate court must always decline to fix trial costs. Its reasoning is grounded in the particular balance of success, the abandoned issue, the difficulty of separating costs, and the procedural concern about factual findings.

How later courts have treated it

This judgment does not discuss any later judicial treatment of itself. It was delivered on 25 May 2026, and the supplied text contains no subsequent cases applying, distinguishing, or otherwise considering this decision.

The authorities it does refer to are earlier decisions. It cites the earlier appeal in M Johnson Building Pty Ltd v O’Toole [2026] ACTCA 8 for the background and partial success on the double-recovery issue ([1]). It also refers to Wu v Wu (No 3) [2024] ACTCA 35 at [8] as the source of general principles concerning costs ([10]). The text does not describe a line of subsequent cases or claim that the present decision has established a broader precedent beyond the reasoning set out in its paragraphs.

Still-open questions

The judgment leaves the amount of first-instance costs unresolved. The parties had not agreed on the process and quantum, and the appellate court declined to determine them in this proceeding ([16]–[17]). It also does not state the ultimate amount of the credit to be given for the Certifier settlement. The earlier appeal had required credit for part of the settlement, but the calculation remained to be worked out.

The reasons do not explain how the costs of each individual appeal issue were calculated, or how much of the total work related to the abandoned wet-areas issue. They say only that the costs were significant enough to consider and that the issues were not sufficiently discrete for particular time periods to be carved out ([11]–[12]).

The judgment also does not provide a general threshold for when issues are sufficiently discrete to permit issue-based costs allocation. That assessment is left dependent on the circumstances and the material before the court. Nor does the judgment formulate a comprehensive rule for costs where each side has partial success. It records a fairness-based outcome in this appeal.

Finally, the judgment identifies possible ways to resolve first-instance costs, including an application in the original proceeding, assessment, or agreement, but does not decide which process would ultimately be used or what findings it would produce ([17]).

This is general legal information, not legal advice. Zoe can make mistakes. Before acting on or relying on this analysis, have it checked by a qualified Australian lawyer.