Recent Full Federal Court decisions have sharpened focus on the “best method” requirement in Australian patent law, particularly in the context of divisional applications. While often overlooked, this requirement can have significant implications for validity and is increasingly being tested in contentious proceedings. Read on to understand what’s changed and how it impacts patent strategy.
Across the globe, an assessment of the validity of a patent/application will usually require a comparison between the invention defined in the claims and the prior art. In this way, the novelty and obviousness of the claims may be determined. Additional requirements, typically support and enablement, will also be considered.
A further requirement that is peculiar to relatively few countries, such as Australia and New Zealand, is the disclosure in a patent application of the best method known to the applicant as of the filing date of the application. This is a serious requirement as it is a ground for patent revocation and has seen renewed prominence as a ground used in invalidity proceeding such as oppositions before the Australian Patent Office, and revocation matters within the Australian Federal Courts.
In our previous article, Dometic discharged? Federal Court clarifies the relevant date for assessing best method, we reviewed a first instance decision where the Court was required to determine whether the date of assessment of disclosure of the best method was:
Section 40(2)(aa) of the Australian Patents Act 1990 requires that a complete specification must disclose the best method known to the applicant of performing the invention. This disclosure is over and above the sufficiency and enablement requirements of the Patents Act 1990. The intention is to provide the public with the full benefit of an invention once the relevant patent rights cease.
In The NOCO Company v Brown and Watson International Pty Ltd [2025] FCA 887 (NOCO), the Court found that all relevant claims of NOCO’s patents were invalid for a lack of inventive step. But since Brown and Watson International Pty Ltd had challenged validity on the grounds of lack of disclosure of best method, the Court opined on this ground. In essence, the Court found that:
In The NOCO Company v Brown and Watson International Pty Ltd [2026] FCAFC 44, the Full Federal Court of Australia found unanimously that the correct date for assessing the disclosure of the best method was the date of filing of each of the divisional patent applications. Consequently, the relevant claims were deemed invalid because the best method had not been disclosed in each of the divisional patents.
In reaching this conclusion at [229], the Full Court noted that Section 40(2)(aa) requires a complete specification to disclose the best method. Further, a divisional application according to section 79B(1) is a complete application and the Dictionary to the Patents Act 1990 defines a complete specification as a specification filed in respect to a complete application. Therefore, Section 40(2)(aa) must refer to each specification filed starting with a PCT application, which is defined as a complete application, as well as each complete specification filed in respect to each divisional application.
The accuracy of this conclusion was reinforced by reference to Section 138, which deals with the revocation of a patent. Specifically, if an application for revocation of a patent is made under this section, the complete specification to be considered is that filed in respect of that patent. It is not a specification filed in respect of the parent PCT.
Nevertheless, divisional practice remains attractive because divisional applications can be “daisy chained”. Double patenting is not an issue unless claims are identical in scope. Divisional applications may be filed during pendency or within 3 months of publication of acceptance (allowance) of a parent application.
But in view of the full court decision, for every divisional patent application filed, applicants will need to carefully review the need to determine the necessity of including an updated best method. This will be required even if the purpose of the divisional is to continue prosecution of an abandoned parent.
Similarly, if the purpose of the divisional is to protect a second or subsequent invention disclosed in a parent application, the same requirement will apply. That is, if a best method for the second or subsequent invention was disclosed in a parent application, it seems unlikely that updates are required. However, if a best method for the second or subsequent invention was not disclosed in a parent application, then it will need to be included in the divisional specification.
From a policy perspective, this appears especially true where a patentee expands their monopoly by including new inventions not disclosed in the parent. As the Full Court noted at [222], to “not to require such disclosure would be at odds with the policy underpinning s 40(2)(aa) and would allow the patentee to withhold knowledge of the best method known at the later time”.
If the purpose of the divisional is to protect an invention not disclosed in the parent, as was the case with NOCO, a divisional application may be inappropriate. As the Full Court noted at [220], to attain divisional status only requires one claim to be entitled to priority from the parent.
“Section 79B does not preclude the divisional application also including other claims that are not clearly disclosed in the parent, although the price to pay for including these other claims would be a later priority date.”
This later priority date would usually be the divisional filing date which often occurs many years after the publication of any originating parent/PCT application from which it is based. That being the case, it is likely that the parent/PCT application would be citable prior art as its publication would likely falling outside Australia’s 12-month grace period. Presumably not novelty destroying but citable for inventive step.
In such circumstances, it would seem preferable to file a patent of addition in place of a divisional. This would have the same priority date as a divisional, i.e., the filing date of the patent of addition application. But the parent whilst citable for novelty is excluded as an inventive step citation.
Numerically few patents/applications have been invalidated because of a lack of best method. NOCO was unusual because of the clear evidence of a commercial embodiment that was “better” than that disclosed in the PCT. Certainly in prosecution, it will be rare for an examiner to gain awareness of a “better method”. Even in pre-grant opposition proceedings, it is unlikely to be an issue.
However, once in a court proceeding and with the benefit of discovery, it is more likely to arise. Still, there is the vexed question of deciding what constitutes “best” and then determining if and when the patentee was aware of this best method. In Dometic Australia Pty Ltd v Houghton Leisure Products Pty Ltd [2018] FCA 1573, the court found that although an inventor was aware of a “better method”, that was insufficient to revoke the patent because the patentee was not aware of this “better method”.
Finally, an appeal of this decision may be made to the High Court of Australia. This requires special leave of the Court, which having regard to the clarity and reasoning of the Full Court, is unlikely to be given.