Complete List of HTML Entities

HTML entities explained

What is a character entity reference?

An HTML entity is a piece of text that starts with an ampersand (&) and ends with a semicolon (;) and that the browser replaces with a single character when it renders the page. Entities exist for two reasons: to write characters that have a special meaning in HTML markup (<, &…) and, historically, to write characters that the document's encoding or the author's keyboard could not produce (©, é, →).

There are three ways to write the same character:

Numeric references always refer to Unicode code points (ISO 10646), whatever the encoding of the page. 😀 is 😀 even in an ISO-8859-1 document. Code points that are not valid characters (surrogates, most control characters, values above 10FFFF) are errors; browsers replace them with the replacement character U+FFFD, and the C1 range €–Ÿ is remapped to the windows-1252 characters for compatibility with old pages.

When do you need entities?

Always escape the characters that HTML itself uses, otherwise the browser will interpret them as markup — this is also the basis of protection against cross-site scripting (XSS) when you display user input:

Everything else is optional. With a UTF-8 document (<meta charset="utf-8"> — the default and the only encoding you should use today), you can type ©, €, é, → or emoji directly in the source; entities only make the file harder to read and slightly larger. Entities remain useful for:

• Characters that are invisible or ambiguous in an editor: the non-breaking space  , the soft hyphen ­, zero-width joiners ‍, directional marks ‎ / ‏, thin spaces  .
• Documents served in a legacy single-byte encoding (ISO-8859-1, windows-1252) that cannot represent the character at all.
• Source code that travels through tools which might re-encode it incorrectly (e-mail templates, old CMS databases in latin1).
• Showing literal markup in a page, e.g. <div> in a tutorial. Use the HTML Escape / Unescape tool to convert whole snippets.

Named vs numeric references

• Named references are readable (© is easier to recognise than ©) but only exist for a fixed list of characters: 252 in HTML 4.01, 2,231 in HTML5 (which added things like ✓, ☆, 
 and hundreds of mathematical symbols).
• Numeric references cover all of Unicode and are the only option for characters without a name (most emoji, CJK ideographs, 😀).
• In XML (including XHTML, SVG and RSS feeds) only five names are predefined — amp, lt, gt, quot, apos.   or © in an XML file are errors unless a DTD declares them, so use numeric references ( , ©) there.
• Inside <script> and <style> elements entities are not decoded: < in JavaScript is the literal four characters. JavaScript strings use \u00A9 escapes, CSS uses \00A9.
• HTML5 also tolerates some legacy references without the trailing semicolon (&copy, &amp, &lt), which is why a URL such as ?x=1&copy=2 can unexpectedly render ©. Always write the semicolon and always encode & as & in attribute values.

The non-breaking space ( )

  (U+00A0) is a space that prevents a line break and is not collapsed with adjacent whitespace. It is the most used — and most misused — entity:

• Good uses: keeping a unit attached to its number (10 km, 25 °C), keeping initials or titles with a name (Dr. Smith), French typography before : ; ? !, and preventing a widow word at the end of a paragraph.
• Bad uses: indenting text or adding horizontal space between words — use CSS (margin, padding, letter-spacing, white-space) instead. A row of   breaks responsive layouts and screen readers.
• Empty table cells and elements are often filled with   to give them a height; the CSS empty-cells: show and min-height are cleaner alternatives.
• Note that trim() in JavaScript and PHP trim() handle U+00A0 differently: JavaScript's String.prototype.trim() removes it, PHP's trim() does not (it only strips ASCII whitespace).

Decoding entities in code

// JavaScript (browser): let the parser decode
const txt = document.createElement('textarea');
txt.innerHTML = '€ 5 & ©';
console.log(txt.value);                       // "€ 5 & ©"

// PHP
echo html_entity_decode('€ 5 & ©', ENT_QUOTES | ENT_HTML5, 'UTF-8');
echo htmlspecialchars('<b>"Tom" & Jerry</b>', ENT_QUOTES, 'UTF-8');  // encode the 5 reserved characters

// Python
import html
html.unescape('&euro; 5 &amp; &#169;')       # '€ 5 & ©'
html.escape('<b>"Tom" & Jerry</b>')

Never build your own decoder with a hand-written list of a few names: use the platform function, which knows the full HTML5 table, the legacy no-semicolon forms and the numeric syntax.