How to Choose a Baby Name: A Practical Order From the Surname to One Name

A long list of favorites is not a decision. The useful order is short: say the first name with the surname, keep a meaning only when the name page states one, read several years of counts, then cut the list until one name is left. The middle-name guide is the next pass, after the first name is real.

Every count below is from the U.S. Social Security Administration file we publish, 1940 to 2025. A rank is within one sex. Counts under five babies per sex per year are absent, so a missing number is "under five," and a true total can be a little higher.

Say it with the surname first

The name you will use is the whole string. A first name can be lovely alone and clumsy in front of the surname you already have.

Our full name tester and the name generator use the same checks. A first name and a surname with different syllable counts are treated as balance. Two one-syllable parts can sound clipped. If the first name ends with the same consonant the surname starts with, the tester says they run together. If both of those letters are vowels, it says the join can blur when you hurry. The last three letters of each, when they match, are a rhyme. The generator's own example of a run-together is Jack Kelly: Jack ends in k, and Kelly starts with k.

The same rules, on names in our database:

• Liam Miller is two syllables plus two. Liam ends in m, and Miller starts with m, so the tester flags a run-together.
• Emma Anderson is two plus three, which counts as balance, and a vowel-to-vowel join, which the tester flags as a blur.
• Theodore Miller is three plus two, with a clean break between the e and the m.
• Olivia Smith is four plus one: a clean break, and the length difference the tester treats as balance after a long first name.

Say the pair out loud anyway. The join check is the last letter and the first letter, which catches Jack Kelly and misses some sounds a recording would catch. Initials, the classic monogram, and a middle name are a separate pass. James Smith is the worked example in the middle-name guide: one syllable plus one, and a shared s. James's origin is debated, so this article gives James no meaning. In 2025 James ranked #6 for boys, with 11,945 boys and 59 girls.

The tester also compares the surname with the 2010 U.S. Census. The Full Name Report's shared-name figure multiplies the latest first-name count by that surname's share of the 2010 population. It assumes the two are independent, which they are not exactly, so treat it as a ballpark.

Print a meaning only when the page states one

A meaning is safe to put on a card when our data marks it sourced. The short line and the etymology note do not always use the same words. Print the line you would actually say, and read the note before you shorten it.

Olivia shows "olive tree." The note says Latin, "olive; olive branch." Liam shows "strong-willed protector (short form of William)." The note says it is derived from William, Proto-Germanic "will" + "helmet." William's own short line is "resolute protector," and William's note is the same "will" + "helmet." The two short lines are not the same sentence. William's boys' high is 67,002 in 1947.

Theodore shows "gift of God." The note says Ancient Greek, "god" + "gift." Emma shows "whole, universal." The note says Frankish, "entire." Michael shows "who is like God?" and the note uses that same question, from Biblical Hebrew. Michael's boys' high is 92,781 in 1957.

Four names in or near the 2025 top ten do not get a meaning here. The page marks the origin debated and keeps a theory beside it.

Charlotte ranked #2 for girls in 2025, with 13,400 girls and 8 boys. The theory is "free (feminine of Charles)." 2025 is the girls' high. Noah ranked #2 for boys, with 20,358 boys and 468 girls. The theory is "rest, comfort." Oliver ranked #3 for boys, with 14,939 boys and 24 girls. The theory is "olive tree." Olivia's page states that gloss. Oliver's page does not. Eliana ranked #10 for girls, with 8,191 girls and 5 boys. The theory is "my God has answered." 2025 is the girls' high.

Jennifer is the long arc of a debated name. The theory is "fair, white, smooth (form of Guinevere)." The girls' high is 63,600 in 1972. In 2025 Jennifer ranked #586, with 501 girls. Boys were under five.

Read the curve, not one rank

A single rank hides the direction. Rank and count can move in opposite directions, because a rank is a place in line and a count is a number of babies.

These are boys' counts. The 2015 rank is the rank in that year's file. The other sex is in the paragraphs above when it is large enough to publish.

Liam's high is 22,252 boys in 2024, so the 2025 count is a little under the peak and still first. Those 20,818 boys were about 1 in 82 of the 1,708,076 boys in the 2025 file. With the 28 girls, 20,846 babies were about 1 in 159 of the 3,315,343 babies in the file.

Noah's boys' count rose from 2015 to 2025, and the rank fell from #1 to #2 because Liam passed it. The boys' high is 20,409 in 2024. The 468 girls in 2025 are part of the name. The rank on the page follows the sex with more babies.

Theodore's 2025 count is the high in the file, up from rank #99 in 2015. William and Michael moved down and stayed familiar, under the 1947 and 1957 highs. Aiden's boys' high is 16,030 in 2009, and the rank has since fallen from #13 to #62.

Girls' counts, same years:

Olivia's count fell while the rank rose from #2 to #1. The high is 19,840 girls in 2014. In 2025 those 13,544 girls were about 1 in 119 of the 1,607,267 girls in the file. With the 10 boys, 13,554 babies were about 1 in 245 of the 3,315,343. A number-one name can be a smaller crowd than it was ten years ago.

Emma was #1 in 2015 and #3 in 2025. The girls' high is 22,719 in 2003. Charlotte's 2025 count is the high, and so is Eliana's, after a climb from #102. Jennifer's 501 girls in 2025 were about 1 in 3,208 of the girls in the file, against a 1972 high of 63,600.

Year-to-year jumps are in the fastest-rising names of 2025. Each name page charts the peak.

Add a second spelling only when you mean the sound

The Social Security file ranks spellings separately. Two spellings are two names. Add the counts when you care how often the sound is used. Keep them apart when you care how often this exact spelling is used.

Aiden means "little fire." The note calls it a variant of Aidan, Scottish Gaelic, literally "fire." Aidan's short line is "fire," and the note is the same word. In 2025 Aiden had 5,046 boys and 27 girls, rank #62. Aidan had 1,043 boys and 8 girls, rank #326. The boys' counts add to 6,089. That sum is not a rank. Aidan's boys' high is 10,067 in 2003, and by 2015 the rank was #185.

Sophia and Sofia both show "wisdom." Sophia's note says Ancient Greek, "wisdom." Sofia's note says "wisdom, especially divine wisdom." In 2025 Sophia had 12,561 girls and 11 boys, rank #5. Sofia had 8,252 girls and 9 boys, rank #9. The girls' counts add to 20,813. In 2015 they added to 27,156 (Sophia 17,454 at rank #3, Sofia 9,702 at rank #14). Sophia's girls' high is 22,336 in 2012. Sofia's is 9,702 in 2015.

If the card will say "wisdom," either spelling can carry that short line. If the card will quote the note, Sofia's note has three extra words.

Cut the list in three passes

A list of twelve favorites does not get shorter by staring at it. Each pass should remove names.

1. Run the surname through the generator. The name generator scores our list against style, origin, the surname, and siblings, and it returns ten names. It is free. Use it to see which of your ideas the checks already flag, and to meet names you had not typed.
2. Make two people choose. A partner match uses one deck. Each person likes or passes in private, and only a mutual like is revealed. A name bracket takes 4, 8, or 16 names and asks for a winner two at a time. Use the match when you want the overlap. Use the bracket when you want a single winner.
3. Give other people one link, then stop. A shortlist lets family vote, thumbs up or down or a rank, with no account. Read the comments. Then put the final two, with the middle name, into the full name tester. The middle-name guide covers rhythm and initials, which this pass needs and this article does not repeat.

A vote total is a preference. It does not fix a join the tester flags, and it does not turn a debated origin into a meaning you can print.

Put that one name on the paper you will send

Once the name is the name, the next job is the paperwork and the announcement, in that order.

States set the birth-certificate deadline themselves. Ask your state vital records office before you promise relatives a spelling you might still change. The due date calculator is a window for when you will want the name ready. It is not a birth date.

An announcement needs the name and the date. Weight, length, and a meaning are optional. The wording guide has lines you can adapt. Add a gloss only when the page states one: "Olivia, olive tree" matches the short line. "Charlotte, free" does not, because that origin is debated. Liam's card can say the short line, or it can say the note's "will" and "helmet." Those are different sentences.

The Baby Kit puts the chosen name on the invitations, the announcements, and the wall art in one style. The meaning line belongs on the wall only when you would say it out loud.

Sources

• Meanings: our name database. Etymology notes quoted above are adapted from Wiktionary (CC BY-SA 4.0). Each name page links its source, and our sources page explains the method. Debated origins are labeled on the name page and are not stated as fact here.
• Popularity: U.S. Social Security Administration baby name data, 1940 to 2025. Ranks are within one sex. The 2015 rank is that year's rank for the same sex. Counts under five babies per sex per year are absent from the file. The 2025 file total is the sum of published counts that year, 1,708,076 boys and 1,607,267 girls. "About 1 in" figures divide that file by the count named in the sentence.
• Surnames in the tester: 2010 U.S. Census surname file, as used on our full name tester. The shared-name estimate assumes a first name and a surname are independent.