Portugal's Seasonality Shift: Winter Nights Up 21%, Summer Up 9%, and Two Points of Gap Closed, Mostly Before 2023

Which seasonality number, and where the floor is

INE defines its seasonality rate as the relative weight of the three months of highest demand in the annual total. Nationally those months are July, August and September, so the rate equals the July to September share. The lowest value a country can post is 25%, which is three equal months out of twelve. Madeira's tourism strategy says so on page 20, and it is the right floor for reading the rest of this article. Portugal's 36.4% sits 11.4 points above it.

Other measures in circulation are not interchangeable with that one. Eurostat users quote the July and August share, which was 26.0% in 2025 and 27.5% in 2019. Airport analysts quote peak-to-trough ratios. The Banco de Portugal tracks the share of non-resident nights from June to September, which was 48.4% in 2014, 45.5% in 2019 and 44.1% in 2024. Turismo do Algarve quotes an "época alta" share whose months it does not define. A regional development body, CCDR-Norte, publishes an index that puts Portugal at 27.3% for 2024, against INE's 36.6% for the same year. These numbers measure different things, and a reader who sees 27.3%, 36.4%, 40.8% and 44.1% in four documents is looking at four definitions.

Two scope notes apply to every figure below. Eurostat's monthly series cover hotels, short-stay accommodation and campsites, and INE's monthly releases exclude campsites. In August 2025 INE counted 10.67 million nights where Eurostat counts 12.62 million, and for the full year Eurostat's 89.43 million compares with INE's 89.7 million. Unless a figure is attributed to INE, it comes from Eurostat and the shares are our derivations from monthly rows. Winter means January, February, March, November and December of the same calendar year. The full-year national total in our Portugal tourism statistics 2025 report uses the same Eurostat basis.

The national series: two points in six years

The first table sets the annual record. It shows something the 2025 headline hides. The July to September share fell 2.85 points between 2015 and 2019, about 0.7 points a year, and 2.00 points between 2019 and 2025, about 0.33 a year. Of those two points, 1.44 came by 2023. The August to January ratio shows the same sequence: 4.33 in 2015, 3.62 in 2019, 3.29 in 2023 and 3.21 in 2025. On this evidence Portugal's seasonality was already converging before the pandemic and the convergence has slowed since.

The monthly rows explain where the movement came from. Every month outside the summer peak grew by between 18.8% and 23.1% between 2019 and 2025: January 20.3%, February 23.0%, March 19.4%, April 18.8%, May 19.1%, October 21.0%, November 23.1% and December 20.4%. June grew 11.8%, July 11.5%, September 10.1% and August 6.8%. August carried 14.1% of the year's nights in 2025 against 15.2% in 2019. Easter fell on 21 April in 2019 and 20 April in 2025, so the March and April comparison between those two years is not distorted by the calendar.

Who filled the winter

Residents did, in relative terms, and non-residents did in volume. Residents' winter nights grew 24.2% between 2019 and 2025 while their July to September nights grew 4.7%. Non-residents grew 19.6% in winter and 12.0% in summer. Foreign visitors supplied 62.1% of the extra winter nights, 2.62 million of 4.21 million, and 80.6% of the extra summer nights. Residents account for about a third of all nights but for more than half of the national fall in the July to September share.

Splitting the 2.00-point fall in the national July to September share by group gives about 1.09 points from residents, 0.88 from non-residents and 0.04 from the shift in weight between the two (derived, Eurostat monthly rows). Residents' own July to September share fell from 43.6% to 40.4%, and their August nights grew only 2.3%. Non-residents' share fell from 35.7% to 34.4%. The data do not say why residents moved. Peak-month prices and school and company holiday calendars are plausible reasons, and we have no series that tests them. Our companion piece on Portugal's domestic tourism covers the size of the resident market and where it travels.

Fuller hotels in the shoulder months, but no monthly price series

Hotel occupancy shows the same shape. In Eurostat's monthly hotel series, bedroom occupancy rose most in the shoulder months between 2019 and 2025: 7.2 points in May, 5.6 in April, 5.1 in October, 5.0 in February and 4.8 in March. It rose 2.5 points in August and 2.4 in September. Measured on bed places, August and September fell by 0.8 points, so peak-month hotels are no fuller than before. A January hotel bedroom was still 40.3% occupied in 2025 against 79.1% in August.

Annual INE figures, which cover hotels, local accommodation with ten or more beds and rural tourism, show that prices rose far more than occupancy. Room occupancy went from 55.3% in 2019 to 58.2% in 2025. The average daily rate went from €89.2 to €124.5, up 39.6%, and revenue per available room from €49.4 to €72.4, up 46.6%. These are nominal euros. Nights grew 15.3% over the period, so the gain was not volume alone. No source gives price by month for 2019, so we cannot say whether shoulder-month rates rose faster or slower than August's. Our RevPAR and ADR analysis deflates the annual series and finds that price built most of the real gain.

Region by region

The national average hides regions that behave very differently. Madeira and Greater Lisbon are almost flat through the year: August is 1.64 times January in Madeira and 1.96 times in Greater Lisbon. The Algarve is at the other extreme among large regions, with August 5.26 times January in 2025, and only the Alentejo is more peaked at 5.30. Our regional articles describe each in detail, among them the Algarve, Madeira, Azores and Alentejo reports.

Two facts from the table matter for the headline question. First, the Algarve's winter share has hardly moved. It was 19.9% in 2023, 20.7% in 2024 and 20.2% in 2025, and its July to September share was 41.6%, 41.3% and 41.3%. Second, the Algarve was not where Portugal's growth happened. Its nights in 2025 were 22.94 million against 22.99 million in 2019, a fall of 0.2%, while Norte grew 34.8%, Madeira 33.2% and the Azores 34.0%. Norte, Madeira and the Azores together supplied about 62% of the national growth (derived), and the Algarve's share of national nights fell from 29.6% to 25.7%.

A counterfactual test suggests the national shift is mostly not a change in regional mix. Applying each region's 2025 July to September share to the 2019 regional weights gives a national share of 36.59%, against an actual 2025 value of 36.35% and a 2019 value of 38.35%. Mix explains about 0.24 of the 2.00 points, or 12%, and the remaining 88% comes from regions becoming flatter themselves. The test uses 2025 within-region shares because Eurostat has no regional monthly data before 2020, and the 2019 Greater Lisbon and Setúbal boundaries differ from today's. It is indicative. A proper regional comparison with 2019 would need INE's municipal tables, which our research tools could not reach.

The airports tell a narrower story

Passenger data point the same way at the national level and diverge by airport. Across all Portuguese airports the November to March share rose from 31.6% to 32.3% between 2019 and 2025, and winter passengers grew 24.8% against 22.1% for the year. Faro, the Algarve's gateway, carried 2.03 million passengers in winter 2025 against 1.67 million in 2019, a gain of 21.8%, and its winter share rose from 18.4% to 19.6%. That is still less than half of the 41.7% a perfectly flat profile would give. Winter 2025/26 scheduled seats at Faro were 1.34 million, 5% more than the year before, against about 4.6 million in summer, so summer has 3.4 times the capacity. Part of the winter growth was domestic flying and larger aircraft, not new international routes. Our Faro airport analysis covers the airline and seat detail.

Other gateways moved the opposite way. Lisbon flattened, with August over January falling from 1.58 to 1.45. Porto and Funchal became more peaked, with the ratio rising from 1.64 to 1.86 at Porto and from 1.41 to 1.65 at Funchal, because summer grew faster than winter. Ponta Delgada went from 2.40 to 2.67 and its winter share fell from 28.5% to 26.5%. For the islands, see our Madeira airport and Azores air access articles, and for the Porto picture the Porto airport report. A city airport with year-round business and leisure demand is a poor proxy for resort seasonality, so these ratios say more about Lisbon and Porto than about the Algarve.

Source markets differ even more. Eurostat's partner-country series, which count direct-flight passengers between Portuguese airports and each country, show Poland moving from a summer market to a year-round one. Its winter share of passengers rose from 21.0% to 38.2% between 2019 and 2025 and its peak-to-trough ratio fell from 4.23 to 1.47. Ireland's winter share rose from 17.7% to 23.6%. Germany (32.3% to 31.7%), Spain (35.4% to 33.1%) and France (33.6% to 33.7%) did not move toward winter, although these totals include Lisbon and Porto city traffic. The British market, the largest winter market in the Algarve, cannot be tracked this way because Eurostat has published no UK partner series since 2020. In the first quarter of 2026, UK residents made up 15.6% of non-resident nights nationally, and the Algarve took 41.2% of their nights.

Spain, Italy, Greece and Croatia

Portugal's change is large against its neighbours, and its level is still close to Spain's. Both countries now have a July to September share near 36 to 38% and a peak-to-trough ratio near 3.2. Portugal's share fell 2.0 points between 2019 and 2025 and Spain's 1.3. Greece did not move, at 56.3% and 56.5%. Croatia's share fell 2.8 points, but its busiest month still carries 41 times its quietest. Portugal's winter share of 27.0% compares with 25.6% in Spain and 6.9% in Greece. The Banco de Portugal's June to September measure for non-resident nights tells the same story for 2024, with Portugal at 44.1% against Spain 46.7%, Italy 54.9%, France 55.5% and Greece 74.9%, as reported by ECO in October 2025 and not independently re-fetched here.

The wider comparison is in our Southern Europe benchmark and in Portugal vs Greece, which uses the same monthly series.

What the targets say, and the claims to drop

Portugal has had a numeric national seasonality target since 2017, although our own earlier notes said it did not. Resolution of the Council of Ministers 134/2017, which adopted the Estratégia Turismo 2027, sets the goal to "reduzir o índice de sazonalidade de 37,5% para 33,5%". The resolution prints no base year, no target year and no definition. Elsewhere the same text says the index fell from 37% to 35% in 2016, and a 2021 hotel and restaurant association plan quotes 36.3% as the starting point for the same 33.5% goal. If INE's rate was 36.6% or higher in every year from 2014 to 2023, as the phrase "lowest since 2013" implies, then a value of 35% for 2016 cannot be INE's index. The Turismo de Portugal target page gives only the wording, "o índice de sazonalidade mais baixo de sempre", with no figure.

On INE's series the gap to 33.5% is 2.9 points, and the 2025 improvement was 0.2 points. The strategy's horizon is 2027. The successor strategy for 2035 has been approved and awaits public presentation, according to the Secretary of State at an industry congress on 22 October 2026, who spoke of almost 50 million international tourists and more than €53 billion in receipts. Press accounts of that speech and of the strategy's December 2025 launch mention no seasonality target. Our tourism strategy 2035 article tracks what is and is not public. Turismo de Portugal's 2025 activity plan contains no use of the word "sazonalidade". Its first operational priority is to spread tourism across the territory and through the year, with a budget of €13.81 million against €121.88 million for promotion overall, 11.3% (derived).

The campaigns on record are small or unquantified. "Não procures mais longe", launched on 19 May 2026, targets domestic travellers and was framed as storm recovery. Its budget is undisclosed. The Algarve's "É sempre Algarve" campaign aims to reduce seasonality and also has no published budget. The eight claims in the next table are the ones that appear most often in articles and speeches about this subject.

The Algarve claims deserve a closer look because they come from the tourism board and recur in our own earlier work. Barlavento wrote on 16 January 2026 that the region went from 43% in 2014 to 38% in 2024 "com a meta de atingir 37% em 2028". No speaker is named, and four accounts of the PMETA 2028 marketing plan list a seasonality indicator without a number. The board's president gave 38.1% of guests and 40.8% of nights as the 2025 high-season shares on 2 February 2026 and called them the lowest ever, without defining the months. Meanwhile Eurostat's July to September share for the Algarve was 41.3% in both 2024 and 2025. The measures differ, and none is a published series.

2026: the pattern held and did not extend

The year began with a clear break in the Algarve. Algarve nights fell 4.7% in January 2026 and 0.7% over January and February, while Faro airport passengers fell about 1.3% over the same two months, on 3.2% more seats (Eurostat passengers and Cirium seats, as reported by Postal). Turismo do Algarve blamed unusually bad weather, and INE counted 90 municipalities hit by intense weather in January and February, with nights down 1.8% there against up 1.7% elsewhere. March and April rebounded, with Faro passengers up 6.8% and 3.9%, but Easter fell on 5 April in 2026 against 20 April in 2025, so those two months should be read together. Registered unemployment in the Algarve rose 58.8% month on month in November 2025, the largest regional increase, and fell 23.3% month on month in April 2026 and 13.0% in June.

The winter ahead has two known changes. Ryanair closes its Berlin base at the end of October 2026, and one projection puts the loss at about 50,000 seats between Germany and Portugal over the winter. Jet2 plans 6.3 million seats for winter 2026/27, up 8.7%, and adds Gatwick to Funchal, with nothing announced for Faro. The EU Entry/Exit System has produced queues at Lisbon, Faro and Porto in 2026, and no source has measured its effect on winter bookings. The border question is covered in our airport capacity analysis for Lisbon. INE's preliminary 2026 annual figures, due around February 2027, will be the first test of whether 2026 ended more or less seasonal than 2025.

What nobody publishes

Several answers the question needs do not exist in public data. There is no regional monthly series before 2020, so the Algarve's 2019 profile cannot be compared with its 2025 one. INE's own seasonality series for 2013 to 2024 is published only as a chart. There are no monthly nights by country of residence, so the claim that the British or German market fills the shoulder can be tested only through airport data, which stop for the UK in 2020. There is no monthly price series by region and no current count of Algarve hotels that close in winter, because owners are not required to report closures. The last estimate is an association's pre-COVID figure of about half. Seasonal employment in tourism has no sourced national series either. We report these as gaps and have filled none of them with estimates.

What this means for travelers, businesses and analysts

Travelers. The shoulder season is busier than it was, and winter is still the quiet time in hotels. A Portuguese hotel bedroom was 40.3% occupied in January 2025, 49.0% in February and 55.6% in March, against 79.1% in August, and Madeira and Lisbon have the flattest profiles of any region. The Algarve is the region where timing matters most. Its tourism board describes the full season as March to October, and January 2026 was weak. The month-by-month detail is in our best time to visit Portugal guide, and the Algarve choices are in the Algarve travel guide.

Tourism businesses and investors. The structural read is that the gain came in two steps, a steady fall before 2019 and a further fall to 2023, and that the Algarve has not joined it. Its winter share has sat at 20% for three years and its nights are level with 2019. An operator underwriting winter demand on the 2015 to 2023 pace is assuming something the 2023 to 2026 data do not show. Winter seats at Faro are under a third of summer seats, and the January 2026 episode shows how quickly weather can erase a small gain. Thresholds that would change this read: a national July to September share below 36.0% on the INE basis, an Algarve winter share above 21%, or an August to January ratio below 5.0 in the Algarve. The unemployment spike each November, 58.8% in 2025, shows who carries the cost of the remaining gap. See also our labour shortage analysis.

Journalists and analysts. The citable figures are INE's 36.4% rate for 2025 and 36.6% for 2024 (Estatísticas do Turismo, 9 July 2026), Eurostat's 38.35% for 2019 on the same measure, the 35.6% and 23.2% shares of 2019 to 2025 growth that fell in winter and summer, and the 33.5% target in RCM 134/2017. Do not repeat that seasonality is the lowest ever, that the Algarve has a 37% target for 2028, that Portugal has 300 days of sun, or that the Algarve season runs to November. Always name the index, because INE, Eurostat, the Banco de Portugal and the regional boards use five. The finding this piece adds to the series is that the convergence was faster before 2019 than after it, that residents account for more than half of it, and that it has barely touched the one region where it matters most.