How to read IBGE's data on names

Where the numbers come from

IBGE published the names and surnames from the 2022 Census in its tool Nomes no Brasil, with a Technical Note (01/2025, dated November 4, 2025) that explains the method. The base is the list of residents of each household on August 1, 2022, the census reference date, and it covers the 27 Federative Units and the 5,570 municipalities: 203 million people in 90.7 million households. Residents who, on that date, were present or absent for at most 12 months are included. The numbers on this site were taken from the compilation IBGE Nomes no Brasil 2022, made by Gabriel Lobo from the public API that feeds IBGE's tool and published under the CC BY 4.0 license. According to the compilation itself, no data was inferred or added from outside sources: the names have no accents, following the source's standard. We thank the author for this work, which made the data usable in bulk.

How the name was collected and counted

1. 1

   1. Collection

   The enumerator wrote down in the “name” field each resident's first name or compound name and, in the “surname” field, the remaining surnames.

2. 2

   2. Cleaning

   One-letter words, two-letter words made of the same letter twice, and the connectors “da”, “das”, “de”, “des”, “do”, “dos”, “di”, “dis”, “du” and “dus” were excluded.

3. 3

   3. First name

   The first name was split off at the first blank space, and only it was considered in the release.

4. 4

   4. Spelling

   Each spelling was counted as written down (Luis and Luiz are different rows), without accents, tilde or cedilla.

5. 5

   5. Sex

   Sex reflects only the information declared at the time of the interview.

What each number on the page means

Generations, not births

The census is a snapshot of the population on one day. The chart bars show how many people born in each decade were still alive and living in the country in 2022. For that reason, the older bars tend to be smaller even for names that were very common at the time. The chart shows well which generations a name is concentrated in, but it is not a series of how many babies received the name each year. Anyone who wants to know the names given to newborns needs to look at the *registro civil* (civil registry). The Portal da Transparência do Registro Civil publishes the most common names registered in Brazil, by period, region and state. These are data of a different nature: they count births registered at the *cartórios* (civil registry offices), not people enumerated in a census.

Why a bar is missing: statistical confidentiality

IBGE only presents a name when its frequency is at least 20 people in Brazil. In a state, the requirement is at least 15 people with the same name; in a municipality, at least 10. The same filters apply to each decade. When a count falls below the threshold, it is simply not published. For this reason, the sum of the states may not match the total for Brazil, and the sum of the municipalities may not match the total for a state. An example: Kauã has no bars before the 1970s. This does not mean that no one born earlier has this name; it means that, if there were any, there were fewer than 20 in each decade. On the pages, these cases appear as “undisclosed data”, never as zero.

“Proportionally most common”

São Paulo is the most populous state and, for that reason, usually has the largest absolute number of people with almost any name. To compare states, the page uses the share of each state's population that has the name. The case of Maria shows the difference: São Paulo has 2,131,207 people named Maria, the largest number in the country, but, proportionally, the name is most common in Ceará, where Maria is the first name of 12.634% of the population, according to IBGE.

The comparison with 2010

The pages also show the total for each spelling in the 2010 Population Census, which counted the people alive in 2010. The 2022 Technical Note warns that there may be differences between the names collected in 2010 and in 2022 because of the way they were collected: variants are counted as written down, accents are not recorded and sex is the one declared in the interview. A difference between the two totals may reflect this, in addition to real changes in the population. The methodology explains where each dataset comes from.

Frequently asked questions