The EU General Court has dismissed an action challenging the partial cancellation of a ‘Waterdrop’ EU word mark that its proprietor filed four months after the non-use grace period on its own earlier, identical mark had expired. The case is T-541/25, Qingdao Ecopure Filter Co. Ltd v EUIPO, with Waterdrop Microdrink GmbH as the other party; it was reported by Marcel Pemsel for The IPKat on 5 November 2026. According to the notice of the action published in the EU Official Journal, the proceedings were brought on 5 August 2025 against the decision of EUIPO’s Fourth Board of Appeal of 22 May 2025 in Case R 1696/2024-4, and concerned EU trade mark No 18 492 910.
The filing history is the whole case. As The IPKat sets out, Qingdao Ecopure Filter (QEF) and its predecessor built a stack of Waterdrop rights in class 11: an EU figurative mark in 2015 for water filters and air filtering installations; an EU word mark in 2020 for water supply apparatus; the 2021 EU word mark now contested, covering a broad class 11 list that included water-filtering goods; and a further EU word mark in 2022. In 2022 QEF began opposing applications by Waterdrop Microdrink GmbH (WMG) in several jurisdictions. WMG answered in 2023 with an invalidity application pleading bad faith — the absolute ground under Article 59(1)(b) EUTMR. EUIPO rejected it at first instance. The Board of Appeal partly reversed, declaring the 2021 mark invalid for the water-filtering goods on the basis that it was a refiling designed to sidestep the obligation to prove genuine use of the 2015 right. The General Court has now dismissed QEF’s appeal, leaving the mark invalid for the goods identical to the earlier registration and intact for the rest — the partial outcome that Article 59(3) EUTMR contemplates.
The Court did not say refiling is unlawful. The IPKat reports that it distinguished MONOPOLY (General Court, 21 April 2021, T-663/19, EU:T:2021:211) and reaffirmed that a repeat filing is not bad faith in itself; what condemned this one was the intention to circumvent the use requirement. The circumstances it weighed were cumulative and, for a docketing team, uncomfortably ordinary: the contested application landed four months after the earlier mark’s grace period ran out; four Waterdrop filings were made inside roughly three years for class 11 goods; the specifications were identical or heavily overlapping; QEF’s product range had not changed and no fast-moving market justified refreshing the register; and in only one of five oppositions did QEF actually rely on the earlier right it was supposedly maintaining.
Every justification QEF offered failed — and how it failed matters more than that it failed
QEF argued the word mark was not identical to the figurative one. The Court held they were the same sign: both consist of ‘Waterdrop’, and the earlier mark’s standard typeface would not register with the average consumer. Converting a lightly stylised logo into a word mark is therefore not a difference you can build a refiling strategy on. QEF argued that a word mark gives more freedom to modernise the brand. The Court treated that plea as inadmissible because it was raised only before the Court — and found no evidence of any intention to modernise in any event. QEF argued that consolidating goods into a single mark cut renewal administration. The Court pointed out the obvious flaw: filing a fresh identical mark while keeping the old registrations alive adds fees rather than removing them. And QEF argued it had been using the earlier mark when it filed. The Court held the applicant’s intention at filing is what counts, irrespective of prior use.
What this changes
The window of suspicion is wider than the grace period. Under Article 18(1) EUTMR the proprietor has five years from registration to put the mark to genuine use, and the mark stays exposed to a fresh five-year non-use window after that. EUIPO’s own published case law on this point — the Wong lo kat Board decision of 20 December 2022, R 2108/2018-2 — involved a refile six days before the grace period expired. T-541/25 involves one four months after it expired. The lesson is that proximity in either direction is now the first thing an invalidity applicant will plot on a timeline, and ‘we filed after the old mark was already vulnerable, so we were not avoiding anything’ is not an answer. If your renewal and use diary flags a mark approaching year five with thin use, treat a new application for the same sign and the same goods as a documented decision, not a routine one. This is squarely a trade mark docketing problem before it is a litigation problem: the grace-period date, the refile date and the gap between them are the evidence.
The burden is formally on the challenger and practically on you. EUIPO’s guidance records that the onus of proving bad faith sits with the cancellation applicant, but that the proprietor is best placed to explain its intentions at the time of filing. T-541/25 shows what happens when that explanation is assembled afterwards. The commercial logic has to exist in contemporaneous form — a dated rebrand brief, a product-launch plan, a specification reviewed against a new range, a decision memo recording why the earlier registration is being replaced rather than renewed. A justification invented for the appeal is worth nothing, and if it is raised for the first time at the General Court it may not even be heard.
Run the argument at EUIPO, not on appeal. The modernisation plea died on admissibility. Whatever your commercial rationale is, it has to be pleaded and evidenced before the Cancellation Division and the Board of Appeal. The General Court reviews the Board’s decision; it is not a second chance to re-plead the facts.
If you refile, change something real — and let the old mark go. The contested mark survived for the goods that were not identical to the earlier specification. That is the shape of a defensible refile: genuinely new goods or services, a genuinely altered sign, or a genuine change of ownership structure. And because the Court specifically rejected the administrative-efficiency story on the ground that both registrations were being kept, surrendering or deliberately letting the superseded registration lapse is now part of the evidence that the refile was housekeeping rather than evasion.
Your enforcement pattern is evidence of your intent. The Court noted QEF relied on the earlier right in only one of five oppositions. Portfolios are read as a whole. Asserting the new registration everywhere while the older identical one sits unused tells a story about why the new one was filed — and it is a story the other side can assemble from public files.
The real answer to a weak use position is a use file. Article 47(2) EUTMR lets an applicant require an opponent to prove genuine use in the five years preceding the contested application’s filing or priority date, and in the absence of that proof the opposition is simply rejected. That is the pressure QEF was trying to engineer around. The durable fix is unglamorous: maintain a dated, per-class use dossier — invoices, packaging, listings, turnover by territory — so the proof-of-use request is answered rather than feared. Where use is genuinely absent in a class, the honest options are to narrow the specification, start using it, or price in the risk. Before filing again, a clearance and status check through trade mark search at least tells you whether the new application adds anything you do not already own.
Nothing in T-541/25 bars repeat filings, and The IPKat’s own reading is that essentially identical marks for identical goods will now struggle without a demonstrated commercial justification. For anyone with a European portfolio built up by successive filings of the same sign, that is a prompt to audit the stack now — while the justification can still be written down honestly rather than reconstructed under cross-examination.