Every name below is a group of spellings that sound identical. Olivia has no homophone spellings in common use, so its total is unchanged — it gets overtaken by names whose supporters were previously counted seperately.

Sophie is not Sophia. It stays exactly where it was, at 63rd. Sophie is not Sophia — it is said differently, so it is counted separately. The rule throughout is strict: same sound, same name. Nothing else qualifies.

These names are not in the ONS top 100 but they would be if spellings were combined. Each one is popular enough — its support is divided across spellings, so no single spelling clears the bar.

Hannah is the starkest case. It sits outside the ONS hundred entirely, yet on a like-for-like basis it is the 48th most popular girls' name in the country — ahead of names that make headlines every year. Its top spelling accounts for less than half its own total.

The corollary is less comfortable. Seven girls' names hold their place in the ONS hundred only because their rivals are divided. Regroup, and they slip below the line: Darcie (102nd), Scarlett (104th), Rosa (104th), Marnie (107th), Alba (111th), Lara (113th) and Frankie (114th). Nothing about these names has changed. The competition stopped being counted twice.

Run the same exercise on the boys and something strange happens: almost no change at the top of the chart. Muhammad stays first. The top twelve stay in almost exactly the same order. Where the girls' podium is rewritten, the boys' is untouched.

Beneath the top 12, though, the boys' chart fragments just as much — it just does so lower down. Seven names climb into the top 100 on merged spellings, and the most striking of them is Rayan, which the ONS does not list in its top 100 at all. Its 715 registrations are split across seven spellings — Rayan, Rayyan, Rayaan, Raiyan, Reyan, Rayen, Rayyaan — and not one of them is common enough to chart alone. Counted as a single name, it is the 81st most popular boys' name in the country. Zayn (799, five spellings) arrives at 69th; Remy, Cody, Mateo, Raphael and Lenny follow it in. Quieter climbs go to Caleb (75th, once Kaleb is counted with it) and Milo (78th, with Mylo). Out go Bobby, Carter, Vincent, Ralph, Stanley and Nathan.

The difference is where the fragmentation bites. Boys' names splinter in the middle of the chart; girls' names splinter at the very top. That is why grouping rewrites the girls' podium and leaves the boys' alone — and it is why the phrase “the most popular girls' name in Britain” is more misleading than the same expression about boys.

One measure of a name's fragmentation: how much of its own total its leading spelling actually holds. The lower the figure, the more thoroughly the name is split against itself.

The mirror image is a small group of names with no homophone variant anywhere in the register — not one alternative spelling that sounds the same, at any level of recorded popularity. Their ONS ranking is exactly what it appears to be, because there is only one way to write the sound.

This is a narrower claim than it sounds, and worth stating carefully. Plenty of top-100 names have variants that merely look related: George has Georgie, Georgi and Jorge on the register, and none of them merges, because none of them is pronounced like George. The names above have nothing at all — no rival spelling of the same sound competing for their count. Florence at 5th and Charlotte at 25th are, in that sense, the most honest numbers in the entire release. Nobody is quietly inflating them, and nobody is quietly draining them away.

The practical lesson is blunt: the chart understates how common some names really are, and choosing a different spelling does not make a name rarer. If you liked Sophia but picked Sofia to avoid the crowd, you did not avoid the crowd — you joined the most popular name in the country by another door. Your daughter will still share her name, out loud, with three thousand others.

Conversely, a name outside the ONS hundred is not necessarily unusual. Hannah, Sara and Amira are all far more common than the published list suggests. If you want a genuinely uncommon name, the number to look at is not the rank of the spelling — it is the rank of the sound.

We hold a pronunciation for every name in our database, written in phonemes. We stripped the ONS 2025 registration counts down to their pronunciations, grouped names whose pronunciations are identical, summed the counts and re-ranked.

The rule is homophones only, and the line is drawn deliberately tightly. Sophia and Sofia merge because they are pronounced identically. Sophia and Sophie do not, because they are not. Nor do Muhammad and Mohammed: they are transliterations of the same name, but they are said differently, and the moment we start merging names that are “really the same” despite sounding different, there is no principled place to stop. Pronunciation is the test, and nothing else is.

A handful of calls are genuinely marginal. Aria and Arya are merged, which lifts Aria to 14th; a reasonable person could dispute it. Where we were unsure how a name is actually said, we left it alone rather than guess. The ONS data itself is unaltered throughout — every count is theirs. All we have changed is what counts as one name.

If you want to see how any individual name has moved over time, every name above links to its full popularity history. And for the year's biggest risers and fallers as the ONS itself ranks them, see our 2025 movers and shakers.