Methodology and sources

Your score

You give each topic a weight from 0 to 5. Your score for a city is the weighted average of its topic scores:

score = Σ (weight × topic score) ÷ Σ weight

• A weight of 0 removes the topic.
• If a city has no data for a topic, that topic counts as average (50) and the city gets a “some data unavailable” label. Leaving it out instead would favor exactly the cities without the data.
• Ties are broken by the overall Social Progress Index.

Topic scores (0 to 100)

Each indicator becomes a 0–100 score by comparing it with the roughly 340 Brazilian municipalities of 100,000 residents or more, which are the cities ranked on this site: the 1st percentile maps to 0 and the 99th percentile to 100, with values beyond those limits capped. Comparing with all 5,571 municipalities, most of them small towns, left many large cities tied at the top of the scale. When lower is better (rent, distance), the scale is inverted. A topic score is the average of its indicator scores.

Rent is the exception: it exists only for the 36 cities, so it is compared among them. Adding a new city can therefore shift the cost scores.

Safety

The Personal Safety component of the Social Progress Index Brazil (IPS Brasil) 2026, built from homicides, killings of young people and women, and traffic deaths recorded by the Ministry of Health (2024 data). We use official data only, not news reports. Health records and state police statistics can differ, so the source and year are shown next to every number.

A second, independent official source: intentional violent deaths (intentional homicides, robbery-murders and bodily injury followed by death) per 100,000 residents in 2025, from the Ministry of Justice (Sinesp), which gathers the state police records. The Safety score is the average of the two indicators. Police records depend on how each state registers occurrences, so small differences between the two sources are expected.

Cost of living

Average asking rent per square meter for apartments (FipeZAP Rental Index, updated monthly). Asking rents describe new contracts, not existing ones. Until republication of exact prices is authorized, the site shows only a band (lower, mid-range or higher rent: thirds of the 36 cities) and links to the source.

Healthcare

Hospital beds (public and private) and doctors per 1,000 residents, from the National Registry of Health Facilities (CNES/DATASUS, September 2026). Doctors are counted as individuals, so someone working in two hospitals counts once. Both are counted where the facility is, so regional hubs score higher than their suburbs. Population: IBGE 2026 estimate.

Jobs

Two indicators from the Central Register of Enterprises (Cempre), IBGE, 2024: formal wage jobs per 1,000 residents, and the average monthly wage of those jobs. It is a picture of the formal job market, not of recent hiring, and it leaves out informal work and self-employment. Jobs are counted where the company is, not where the person lives, so regional hubs rank above their neighbors. The score does not say how easy it is to find work in a given field.

Education

The Ideb, from INEP (Ministry of Education), 2025 edition: the national index of school quality, from 0 to 10, which combines pupils' results on the national assessment (Saeb) with how many pass each year. We use the public-school result (state and municipal systems together) for the early grades (5th grade) and the late grades (9th grade), and the Education score is the average of the two. It covers public schools only, and it is published every two years.

Internet

Two indicators from Anatel, the telecom regulator (August 2026): fixed broadband connections per 100 households, and the share of home connections with a contracted speed of 100 Mbps or more. The first includes business connections, which inflates business districts. We do not use fiber share, which penalizes cities with fast cable internet. Speed-test data (such as Ookla) is not used because its license forbids commercial use.

Airport access

The average of two indicators from ANAC flight records:

• Distance: straight line from the city center to the nearest airport with regular international flights.
• Connectivity: international departures in the last 12 months from all such airports within 100 km. It separates a hub with flights to dozens of countries from an airport with one flight a week, which distance alone treats the same. It uses a logarithmic scale, so each tenfold increase in departures adds the same number of points.

An airport counts as international if it had at least 52 departures to destinations outside Brazil in the last 12 months (about one per week). Nineteen airports qualify.

Climate

Climate is a filter, not part of the score. Monthly averages come from INMET climate normals 1991–2020, or 1981–2010 when the nearby station has no recent normal. Each city uses the nearest station within 100 km whose altitude differs by at most 300 m from the city's, with at least 10 months of temperature data. Bands use the average of daily highs and lows over the year: hot (25 °C or more), warm (22–25 °C) and mild (below 22 °C).

Overall Social Progress Index

Shown on every city page and used to break ties. It summarizes 57 social and environmental indicators (IPS Brasil 2026). The 2024, 2025 and 2026 editions are not directly comparable, so only the latest one is used.

Which cities, and why

All Brazilian municipalities with 100,000 residents or more in the latest IBGE estimate (about 340), the same group used as the scale for the scores. Below that size, indicators per resident swing too much: a single hospital can double a town's beds per 1,000.

Rent is the exception: the FipeZAP Rental Index covers only 36 of these cities (22 state capitals and 14 other cities in São Paulo, Rio de Janeiro and the South), and there is no nationwide source with current rents (the latest census figures are from 2010). The other cities show “no rent data”, and the cost topic counts as average (50) in their score. Use the “only cities with rent data” filter to compare cities on every topic.

Updates and limitations

Sources are updated on different schedules (monthly, yearly or fixed), so each number shows its own reference period. Scores compare cities with each other; they are not absolute measures of quality of life. City descriptions come from the introduction of the Portuguese Wikipedia article (CC BY-SA 4.0), machine-translated into the other languages, and photos from Wikimedia Commons, with author and license on each page; they are context only and do not enter any score. Found an error? Use the “Spotted an error? Tell us” link at the bottom of each city page.

Indicators and sources