Can I just leave the accent off?

Not safely. In Spanish, año means year and ano does not. In Danish, ø is a different letter from o, not a decorated one. Dropping a mark can change the word or make it a misspelling.

Do capital letters keep their accents?

Yes. The Académie française states that accents on capitals are obligatory in French, so it is École and État. Spanish, Portuguese and Czech follow the same rule.

What is the difference between the acute accent and an apostrophe?

The acute accent ´ is a standalone spacing mark that is never used inside a word. The correct apostrophe is ’ and the ASCII fallback is '. To write an accented letter, use the real letter such as é, never e followed by an accent character.

Why does my ß look like a Greek beta?

Because people substitute the wrong character. ß is the German sharp s and β is the Greek letter beta. They are different code points and only one of them belongs in Straße.

Is Romanian ş the same as Turkish ş?

They look similar but are different characters. Turkish uses a cedilla, Romanian uses a comma below. Romanian text should use ș and ț, not ş and ţ.

Common accent mistakes and the confusable characters

Most accent errors are not spelling errors. They are character errors: the right shape from the wrong code point. On screen you cannot tell, but search engines can, sorting can, databases can, and so can any reader who speaks the language. Here are the confusions worth fixing.

The apostrophe family

Four characters that look nearly identical and do completely different jobs.

Character	Code point	What it is for
´	U+00B4	Standalone acute accent. Never belongs inside a word.
’	U+0027	ASCII apostrophe, the typewriter fallback.
’	U+2019	The correct apostrophe and closing single quote.
′	U+2032	Prime: feet, minutes, derivatives.

Use ’ in don’t and it’s. Use ′ in 6′2″ and 45′. Never build an accented letter by typing e followed by ´, because you get two characters that do not combine, look wrong, and break every search.

There is one classic trap here: writing “in the ’90s” with an opening quote. The apostrophe stands for missing digits, so it must be the closing form ’.

The backtick is not a grave accent

The ` key on your keyboard is U+0060, and although Unicode names it GRAVE ACCENT it is the backtick: Markdown code fences, shell command substitution, JavaScript template literals. It never sits over a letter.

To write a grave accented vowel you need the real letter: à, è, ì, ò, ù. In The US International keyboard layout explained the backtick key becomes a dead key that produces exactly those.

Three tildes

~ U+007E is the ASCII tilde: home directories, “approximately”, URL paths.
˜ U+02DC is the small spacing tilde, a dead key clone that sits higher.
The combining tilde U+0303 is what actually builds ñ and ã.

If you type n~ you get two characters, not ñ. And if you mean approximately, ~ is fine but ≈ is better.

ø, ö and ó are three different letters

This one causes real damage in names.

ø is the Danish, Norwegian and Faroese letter o with stroke. ö is o with two dots, used in German, Swedish, Finnish, Hungarian, Turkish, Icelandic and Estonian. ó is o with an acute, used in Spanish, Portuguese, Polish, Czech, Hungarian and Irish.

They sound related in some languages, which is why people swap them. Do not. København and Tromsø take ø, never ö. And ø is not the empty set symbol ∅, and not a slashed zero, however similar your font makes them look.

Turkish ı and i

Turkish has two i letters, and they pair up differently from every other language:

ı (U+0131, dotless) pairs with capital I
i pairs with capital İ (U+0130, dotted capital)

Default Unicode case conversion lowercases I to i and uppercases i to I, which breaks Turkish words every time. The famous case is a 2008 text message in which sıkışınca became a very different word after automatic casing. If you write software that touches Turkish text, use locale aware case functions, and map ı to i explicitly when building URL slugs.

ß is not a beta

ß is the German sharp s, historically a ligature, pronounced as an s. β (U+03B2) is Greek beta. Writing Straβe is wrong in the same way writing Str0sse would be.

The German capital is either SS or ẞ (U+1E9E), which became official in 2017 and is now the preferred form in all caps. It is never a capital B.

While we are here: ß follows long vowels and diphthongs (Straße, groß, heißen) and ss follows short ones (Masse, dass). Switzerland does not use ß at all.

ñ, ń, ň and ņ

Four different letters, four different languages, and they are not interchangeable.

Letter	Code point	Language
ñ	U+00F1	Spanish, Galician, Basque, Filipino
ń	U+0144	Polish
ň	U+0148	Czech, Slovak
ņ	U+0146	Latvian

Writing a Polish koń with a Spanish ñ is the same class of error as writing an English word with a Cyrillic letter that looks the same.

The related mix-up is ç against ҫ. The first is the Latin c with cedilla used in French, Portuguese, Turkish and Catalan. The second is a Cyrillic letter used in Chuvash and Bashkir. They are indistinguishable in many fonts and they break sorting, search and URLs when mixed.

Hyphen, dashes and minus

Four characters, four jobs.

Name	Code point	Use
Hyphen-minus	U+002D	Compounds: well-known, twenty-one
En dash	U+2013	Ranges and relations: 1990 to 1995, Boston to Hartford
Em dash	U+2014	A parenthetical break in a sentence
Minus sign	U+2212	Arithmetic: 5 − 3, −7

The hyphen on your keyboard is doing three of these jobs badly by default. In particular, a negative number written with a hyphen sits at the wrong height and the wrong width next to a plus sign.

Capitals keep their accents

The belief that capitals drop their accents is a hangover from metal type and old typewriters, not a rule.

The Académie française is explicit: accents on capital letters are obligatory. So École, État, À. Spanish keeps the accent on Ñ and on Á in headlines. Portuguese and Czech do the same. If a style guide in your company says otherwise, it is repeating a printing limitation from 1950.

Windows has a separate Alt code for every capital, listed in Alt codes for accented letters: the complete verified list. Case conversion tools should preserve marks too, which is what Case converter does.

Dropping the mark changes the word

The reason all of this matters:

Spanish año (year) against ano. This is the mistake to avoid above all others.
Spanish papá (dad) against papa (potato), sí (yes) against si (if).
French côte (coast) against cote (rating) against coté (rated).
Czech být (to be) against byt (a flat).
Turkish kar (snow) against kâr (profit).

The invisible one: two ways to write the same letter

Finally, the mistake you cannot see. Unicode lets é be one code point (U+00E9) or two (e plus a combining acute). They render identically and compare as different strings, which is why a search for a name sometimes returns nothing. Normalise text to NFC on input, and see HTML entity, Unicode and CSS codes for accented letters for the code detail and What is a diacritic? Accent marks explained for the background.

If a document already carries the wrong characters, Remove accents from text and Add accents to text will normalise it faster than hand editing. To type the right characters in the first place, see How to type accents on Windows (Alt codes and layouts) or How to type accents on a Mac (Option key and press-and-hold), and check the letters your language actually uses at Spanish, German or Turkish, or from Accent Letter.