Hooray for the holiday!

The map also compares regions within the same month using average daily overnight stays per km². This measures accommodation concentration across each whole region. It is not a count of people on streets. Our eventual crowding objective includes residents, commuters and visitors alike.

The score measures a place’s seasonal accommodation activity. Both domestic and international overnight stays are included. It is our calculation from INE figures, not an official crowd rating.¹

For each calendar month of 2025, we divide overnight stays by its day count. Regional figures sum their component subregions before this calculation.

We divide the span from the quietest month to the busiest into five equal bands. Score 1 covers the lowest fifth of that span, and score 5 covers the highest fifth. Months do not have to split evenly between scores.

A score of 1 does not mean empty, and 5 does not prove pavement congestion. Read the percentage of peak alongside the score. Scores compare months within one place. They do not compare crowd levels between places.

Formula: position = (monthly daily activity − annual minimum) / (annual maximum − annual minimum). Score = min(5, 1 + floor(5 × position)). The range uses the twelve monthly daily values for 2025.

Accommodation statistics omit day trippers, many small holiday rentals, visitors staying with friends or family, and residents attending events. The covered accommodation includes hotels, rural tourism, and local accommodation with more than ten beds.²

Weather values are IPMA’s mean daily maximum temperatures for 1991–2020. A warm monthly average can contain colder days and heatwaves. These are planning references for the named stations.³

Business openings need a separate check. Visitor activity cannot establish whether a restaurant, boat service, or seasonal attraction operates. A town’s resident economy and each operator’s calendar matter.

The highlighted months match the controls. If none match all three preferences, months meeting the most criteria appear as alternatives. All twelve months remain visible. Months just outside the range remain candidates. You can adjust the controls.

The shortlist is a screening aid, not a weighted best-month ranking. Rain uses the historical chance of at least 1 mm in a day. The shortlist does not account for daylight, daily visitor patterns, or individual attraction capacity.

• Coimbra: Queima das Fitas brings a May parade through city streets. Check that year’s programme. University visitor services also list academic-event closures, including Latada on 4 October 2026.^{4, 5}
• Porto: São João celebrations take place on the night of 23 June. That evening needs its own crowd judgment.⁶
• Fátima: the major pilgrimage dates include 12–13 May and 12–13 October. A regional monthly average can hide these concentrations.⁷
• Viana do Castelo: Senhora da Agonia takes place around 20 August. Check the annual programme before choosing dates.⁸

Coimbra’s university visitor service publishes winter opening hours for its Botanical Garden, alongside a short list of closing dates for the visitor circuit. This gives evidence of winter operation for those sites, not a guarantee that every local business opens.⁵

For an Algarve attraction example, Zoomarine’s 2026 ticket season runs from 12 March to 28 November. Its operating-day calendar remains the check for a chosen date.⁹

For plans centred on small coastal resorts, treat a score of 1 as a prompt to verify your intended restaurants and activities. For a city visit, check the attractions you want instead of treating low tourism activity as a city-wide shutdown.