New Zealand receives 0.69 visitors per resident and 14 per square kilometre — the lowest of any comparable destination. It has no visitor cap and no access charge anywhere, and one carrying-capacity framework in the whole country whose triggers have never fired. Specific places are busy on specific days. That is a management question.
New Zealand receives 0.69 visitors per resident and 14 visitors per square kilometre — the lowest of any comparable destination we could check, by a wide margin. It has no visitor cap anywhere, no access charge anywhere, and exactly one published carrying-capacity framework in the entire country, whose triggers have never been activated. Specific places are busy on specific days. That is a management question. It is not a national ceiling.
Our ideas and proposals for the future of New Zealand cruise. Forward-looking opinion, not statements of fact.
Two different claims
"New Zealand is full" and "Milford Sound is crowded on a January morning" are not the same statement. They require different evidence and different responses, and conflating them has cost the tourism debate a decade of clarity.
The second claim is often true and always worth taking seriously. The first is a claim about national capacity, and it should be tested against national numbers.
This paper tests it. The finding is that the national claim does not survive contact with the official data — but that the site-level claim is real, under-measured, and being managed with less evidence than it deserves.
Both halves matter. A publication that argues for cruise growth while waving away genuine local pressure will be ignored by the communities whose consent the growth depends on.
Part one: the national picture
New Zealand received 3,674,600 overseas visitor arrivals in the year to June 2026, up 9% on the prior year and 94.5% of the year-to-June-2019 baseline of 3.89 million. The June 2026 month alone reached 95% of June 2019. Christchurch (547,800 arrivals, up 21%) and Queenstown (493,500) both set records.
Set against comparable destinations:
| Population | Land area (km²) | Visitor figure | Per resident | Per km² | |
|---|---|---|---|---|---|
| New Zealand | 5,361,300 | 263,310 | 3,674,600 (YE Jun 2026) | 0.69 | 14.0 |
| Japan | 123,050,000 | 364,500 | 42,683,600 (2025) | 0.35 | 117.1 |
| United Kingdom | 69,487,000 | 241,930 | 42.6m (2024) | 0.61 | 176.1 |
| Croatia | 3,866,233 | 55,960 | 20.2m (2024) | 5.22 | 361.0 |
Land areas from World Bank indicator AG.LND.TOTL.K2, which excludes inland water, applied consistently. Ratios calculated from the two sourced columns.
New Zealand receives roughly one twenty-sixth of Croatia's visitor density per square kilometre, and about one eighth of Britain's.
The caveat that has to travel with that table
We are not going to publish that comparison without the thing that undermines it, because someone else will.
These four countries do not count the same thing.
- Japan counts border entries by foreign nationals. Its definition explicitly includes cruise-ship tourists and transit landings, and includes expatriates and international students re-entering. It counts entries, not people.
- The United Kingdom counts completed visits by overseas residents, estimated from a sample survey. Anyone entering more than once in a period is counted each time. The 2025 annual total does not exist: the ONS says it is "not endorsing comparisons with previous years" following a July 2024 survey redesign, and VisitBritain has actively declined to publish the 2025 quarterly data, stating that it will not upload it to its reports because the volumes look "markedly lower" than other sources suggest. The figure above is 2024.
- Croatia counts registrations in commercial accommodation. A tourist who changes hotel is re-registered and counted again. It excludes day-trippers and cruise passengers entirely — which is a striking exclusion given Dubrovnik's reputation.
- New Zealand counts border arrivals.
So Croatia's number is inflated relative to a border count by re-registration, and deflated by excluding exactly the cruise callers this publication exists to write about. Japan's includes the cruise passengers Croatia excludes.
The comparison is directional, not arithmetic. It establishes that New Zealand is at the low end of visitor density among countries that manage far more, on each country's own official measure. It does not support a precise multiple, and we will not publish one.
That is a weaker claim than "New Zealand has a twenty-sixth of Croatia's tourism." It is also one that survives scrutiny.
Part two: what New Zealanders actually think
The overtourism argument is often made on behalf of residents. It is worth checking with them.
Tourism New Zealand's Views on Tourism survey, run by Angus & Associates with 3,173 respondents weighted to the 2023 Census, for the year to September 2025:
| YE Sep 2025 | Prior year | |
|---|---|---|
| "Tourism is good for New Zealand" | 95% | 95% — stable since 2019 |
| "Tourism is good for my region" | 90% | 88% (a statistically significant rise) |
| "International visitors put too much pressure on New Zealand" | 38% | 36% |
| "International visitors put too much pressure on my region" | 31% | 29% |
| Report at least one adverse local tourism impact | 73% | consistent year on year |
Two readings are available and both are legitimate.
The supportive reading: approval has sat at 95% for six years and regional approval is rising. This is not a picture of collapsing social licence.
The cautionary reading: 73% of New Zealanders report at least one adverse local impact, and the pressure measures are drifting up two points a year in both directions. The top complaints are ordinary and fixable — litter and waste (29%), difficulty parking (24%), damage to the natural environment (22%). Those are infrastructure and management failures, not volume failures, and they are exactly the kind of irritation that hardens into opposition if left alone.
A note on sourcing. "Mood of the Nation," the survey most often cited in New Zealand tourism debate, appears to have ceased after March 2019. The operating national series is Views on Tourism. Anyone citing "Mood of the Nation" for a recent figure is citing something at least seven years old.
Part three: the sites that are genuinely under pressure
Here the argument turns, because the site-level evidence is real — and thinner than it should be.
Milford Sound / Piopiotahi
Visitation went from 437,000 in 2012 to 883,000 in 2018. In December 2025, day-cruise passengers were up 12,117 on the previous December, a 13% rise, with overall visitation running 6% above the pre-Covid peak.
This is the single most-cited pressure point in New Zealand tourism. It is also the one where the government has made a formal decision, and the decision went the other way.
In June 2025 the government committed NZ$15.2 million to Piopiotahi, confirmed that cruise ships continue to access the fjord, retained Milford Aerodrome, and explicitly rejected a previous proposal to ban cruise ship and aircraft access. Further work is committed on infrastructure and long-term planning, with a Special Amenities Area within Fiordland National Park under consideration.
There is no access charge and no visitor cap at Milford Sound.
Tongariro Alpine Crossing
The Crossing is the only site in New Zealand with a published carrying-capacity framework, and it is worth understanding precisely because it is the template for everything else.
| Season | Single-direction | Total, both directions |
|---|---|---|
| 2022-23 | 60,000 | 85,000 |
| 2023-24 | 108,000 | 126,000 |
| 2024-25 | 94,000 | 133,000 |
| 2025-26 | 71,000 | 106,000 |
| Peak (2018-19) | — | ~155,000 |
The 2025-26 fall is attributed by DOC to two fires and unusually poor summer weather, not to management.
DOC's carrying capacity review sets triggers, not limits: consider imposing limits if annual visits exceed 155,000, or if more than five days exceed 2,000 people, or if a rolling five-day average exceeds 1,500. The highest single day recorded was just over 2,000, on 30 March 2024.
No cap has ever been imposed. DOC stated explicitly that there would be no cap on bookings for the 2024-25 season. A booking system was introduced in 2023-24.
Cathedral Cove, Queenstown, and the measurement gap
Mautohe Cathedral Cove reopened in December 2024 after closure from Cyclone Gabrielle, and immediately saw 800-plus people on Boxing Day and again on 2 January. DOC's response was a Trigger Action Response Plan for landslide and rockfall risk, weather and seismic monitoring, and new toilets. No cap, no charge.
Queenstown Lakes District Council projects that on a peak day in 2027 the district will hold 73,375 visitors against 58,700 residents — a total peak-day population of 132,075. By 2057 it projects 142,826 peak-day visitors against 114,261 residents. These are modelled projections, not counts: the methodology sets average-day visitors at 50% of residents and peak-day at 125%. There is no cap or access charge in Queenstown.
And here is the problem underneath all of this. New Zealand's site-level visitor counter data is not published after 2019. The most recent public figures for the Hooker Valley Track (121,000+), Cathedral Cove Walk (37,000+), Roys Peak (75,000+) and Franz Josef Glacier (759,000+) are all for the year ended February 2019. DOC's current public reporting is survey participation and Great Walks bed-nights, not site counters.
So the country is having a national argument about site-level crowding using seven-year-old site-level data.
Part four: the cruise-specific question
None of the above is about cruise. Cruise passengers are a specific case, and the honest position is that they are both a smaller problem and a differently shaped one.
Cruise contributed 1,258,703 passenger visit days in 2024-25 — passenger-days spent ashore, not unique people. Average spend was NZ$313 per passenger per day ashore, up from NZ$282.20 the year before. B2B News reports the comparison against other visitors at NZ$282 versus NZ$324 per day on the earlier year's figures.
The distinctive feature of cruise pressure is that it is concentrated and predictable. A ship arrives at a known hour with a known number of people and leaves the same evening. That is a harder daily peak than independent travel and a far easier planning problem, because the schedule is published years ahead.
It also has a property nothing else in tourism has: it is movable. An itinerary can be redirected to a different port or a different month without asking a single passenger to change their plans. No other part of the visitor economy can be redistributed by an administrative decision.
That is why cruise belongs on the solution side of the dispersal argument, not the problem side — provided the network has somewhere to disperse to, which is a question about port readiness rather than about tourism volume.
Part five: what other places have actually done, and whether it worked
If New Zealand ever needs to limit cruise at a specific location — and on current trajectories that is a long way off — the international record is unusually clear about what survives.
| Place | Measure | Outcome |
|---|---|---|
| Venice | Ships over 25,000 GT, 180m length or 35m freeboard banned from the Giudecca and San Marco canals, from 1 August 2021 | Held. Converted into primary legislation, Law No. 125, on 16 September 2021 |
| Key West | Voter measures (Nov 2020) banning ships over 1,300 passengers and capping cruise visitors at 1,500/day | Nullified. Florida SB 1194 (2021) pre-empted voter-approved maritime-commerce referendums statewide |
| Bar Harbor, Maine | 1,000 disembarkations per day, by referendum, 2022 | Cut back. After four years of litigation, ruled in May 2026 to be constitutional but enforceable only in July and August |
| Juneau | 16,000 lower berths/day, 12,000 Saturdays; five large ships/day | Voluntary agreement. An attempt to put it into an ordinance was postponed 5–4 on 31 August 2026, citing the Bar Harbor litigation |
| Sitka | 7,000 passengers/day | Non-binding memorandum. A binding ballot version was rejected by over 70% in May 2025 |
| Barcelona | Cruise terminals cut from seven to five; new terminal for 7,000 concurrent passengers; €185m | In force. Agreement signed 17 July 2025. No annual cap — concurrent capacity only |
| Amsterdam | 100 sea cruise calls per year from 2026; mandatory shore power from 1 January 2027 | In force. City Council decision, 28 June 2024 |
| Santorini | 8,000 passengers/day, enforced by a ranked berth-allocation algorithm; tightened in 2026 to assume 100% rather than 80% occupancy | In force. Scheduled ships fell from 728 in 2025 to 595 in 2026, down 18.3% |
The pattern is consistent enough to state as a rule.
Hard passenger caps and vessel bans get litigated, pre-empted or narrowed. Key West's were struck out by the state. Bar Harbor's survived but only for two months of the year. Juneau's have deliberately stayed voluntary because of what happened in Maine. Venice is the one that held, and only because it went through a national parliament.
Measures that work through berth allocation, terminal capacity and infrastructure hold. Barcelona, Amsterdam and Santorini all took that route and all three are operating. Santorini's is the most instructive: an algorithm that ranks and allocates berths achieved an 18.3% reduction in scheduled ships without a single ban, and it tightened simply by changing an occupancy assumption.
There is one further detail worth carrying. Juneau's cap is expressed in lower berths, not passengers. Lower berths count two to a cabin; ships routinely sail above that. Juneau ran at 104% of lower-berth capacity in 2024 and 101% in 2023. A "16,000 cap" therefore permits more than 16,000 people ashore. Reporting it as a passenger limit — as this publication has previously done — overstates how tight it is.
Dubrovnik is routinely described as limiting itself to two ships and 4,000 passengers a day. We went to the primary document — the cooperation agreement signed with CLIA on 24 July 2019 — and it contains no numerical limits at all. The figure circulates widely in media coverage and we could not verify it from any primary source. Until someone produces one, we will not repeat it.
Part six: the levy question
New Zealand does have a national visitor charge, and it rose sharply.
The International Visitor Conservation and Tourism Levy is NZ$100 per person, increased from NZ$35 on 1 October 2024. Revenue rose from NZ$62.5 million in 2023-24 to NZ$146 million in 2024-25. Of that, NZ$34.7 million was allocated to tourism projects, NZ$18.74 million to conservation and NZ$3.31 million to administration, with NZ$271.9 million committed forward through 2028-29.
Australian and New Zealand passport holders are exempt, as are most Pacific Island passport holders and transit passengers.
Whether cruise passengers pay the IVL is confirmed to us only from a tourism-marketing FAQ, which states that cruise passengers require an NZeTA and that the levy is bundled into it. We could not confirm this from immigration.govt.nz. For a cruise readership this is not a detail, and we are asking Immigration New Zealand for a definitive answer before we publish anything further on it.
What this adds up to
New Zealand is not full. On the official numbers it is among the least visitor-dense developed destinations in the world, its residents' approval of tourism has been flat at 95% for six years, and it has not imposed a single visitor cap or access charge anywhere in the country.
Some New Zealand places are genuinely busy on some days, and the country is arguing about it with 2019 data.
Those two facts point at the same conclusion. The overtourism debate in New Zealand is not a debate about volume. It is a debate about measurement, timing and dispersal — where visitors go, when they go, and what infrastructure meets them.
Cruise is unusually well suited to that framing, because it is the one part of the visitor economy that can be redirected in space and time by decisions made years in advance. Whether that is an opportunity or an irritant depends entirely on whether the ports have somewhere to send people, and whether anyone is counting.
What we are asking
Restore site-level visitor data. Publishing 2019 counter figures in 2026 is not a defensible basis for national policy. If DOC's counters still run, the numbers should be public.
Publish thresholds before they are needed. Tongariro's trigger framework is good practice and it is the only one in the country. Any port or site that might face pressure should have one — set in advance, in public, with the response attached.
Use berth allocation, not bans. Every jurisdiction that reached for a passenger ban has spent years in court. Every one that worked through berth capacity and terminal design is operating. New Zealand should not have to learn this twice.
Settle the IVL question for cruise. A cruise readership is entitled to know from a primary source whether its passengers pay a NZ$100 levy.
Correct us. Every figure here is sourced and dated, and where a number does not exist we have said so — including for the Dubrovnik limit, the UK's 2025 total and New Zealand's own site-level counts since 2019. If you hold better data, we will publish it and amend the page.
Sources
New Zealand visitors — Stats NZ, International travel: June 2026 and annual arrivals release, 14 August 2026; Minister of Tourism release, 14 August 2026. MBIE discontinued international tourism forecasts; figures circulating for 2026 and 2034 are government targets from the Tourism Growth Roadmap (May 2025), not forecasts.
International comparison — Japan National Tourism Organization, 2025 provisional visitor arrivals (26 January 2026), with JNTO definition notes; ONS Overseas travel and tourism and VisitBritain inbound annual data (26 August 2025); Croatian Bureau of Statistics, Tourist Arrivals and Nights in Commercial Accommodation 2024 (28 February 2025); World Bank land area indicator AG.LND.TOTL.K2.
Populations — Stats NZ national population estimates at 31 March 2026; Statistics Bureau of Japan, 1 October 2025; ONS provisional UK estimate, 30 June 2025 (released 27 November 2025); Croatian Bureau of Statistics mid-2024 estimate.
Sentiment — Tourism New Zealand, Views on Tourism, Angus & Associates, year ended September 2025.
Sites — Department of Conservation: Piopiotahi sustainable destinations pages; Tongariro Alpine Crossing carrying capacity report (June 2024) and monitoring insights; Visitor Sites Data 2018-19; media release on Cathedral Cove, 21 January 2025. New Zealand Government release on Milford Sound, 4 June 2025. Queenstown Lakes District Council Demand Projections, May 2026.
International cruise measures — Library of Congress Global Legal Monitor (Venice); Florida SB 1194 (2021); Sun Journal, 22 May 2026 (Bar Harbor); Juneau Empire, 2 September 2026; Cruise Industry News, May 2025 (Sitka); Port de Barcelona, 17 July 2025; Seatrade Cruise (Amsterdam); TNT Magazine (Santorini and the Greek levy); The Dubrovnik Times, 24 July 2019.
Levy — MBIE International Visitor Conservation and Tourism Levy pages and Annual Performance Report 2024-25 (published June 2026).
Where a figure could not be verified from a published source, this paper says so rather than estimating.