Percent-Encoding Reference: Reserved & Unreserved Characters

What Is Percent-Encoding?

Percent-encoding is the mechanism defined in RFC 3986 for representing data within a URI. It replaces characters with a percent sign % followed by two uppercase hexadecimal digits representing the character's byte value.

Space   → %20
&       → %26
é (UTF-8) → %C3%A9

This reference covers every character you need to know about when building or parsing URLs.

Reserved Characters

Reserved characters have special meaning in URI syntax. When they appear as data rather than delimiters, they must be percent-encoded.

Complete Reserved Characters Table

When to encode reserved characters

• As data in query strings: ?url=https%3A%2F%2Fexample.com
• As data in path segments: /search/hello%20world
• When the character's delimiter role is not intended: ?q=AT%26T (encoding the & in "AT&T")

When NOT to encode reserved characters

• / between path segments: /api/users/42 — leave as-is
• ? starting a query string: ?search=hello — leave as-is
• = assigning query values: ?page=1 — leave as-is
• & separating parameters: ?a=1&b=2 — leave as-is

Unreserved Characters

Unreserved characters never need encoding and should not be encoded. Encoding them produces equivalent but unnecessarily long URLs.

Total: 66 unreserved characters

Why tilde (~) matters

Older specifications (RFC 2396) classified ~ as a reserved character. RFC 3986 moved it to unreserved. Some legacy systems still encode ~ as %7E, but this is unnecessary and should be avoided.

Common Encoded Values Quick Reference

The table below covers frequently encoded characters with practical context:

ASCII Printable Characters Full Table

Every printable ASCII character and its encoding status:

Characters 65–90 (A–Z) and 97–122 (a–z) are all unreserved and omitted from this table for brevity.

Unicode and UTF-8 Encoding

Non-ASCII characters use UTF-8 encoding before percent-encoding. Each UTF-8 byte becomes a separate %XX sequence.

Practical example

// Encoding a URL with Unicode
const name = 'José García';
const encoded = encodeURIComponent(name);
// "Jos%C3%A9%20Garc%C3%ADa"

// The space could also be + in form encoding
const formEncoded = new URLSearchParams({ name }).toString();
// "name=Jos%C3%A9+Garc%C3%ADa"

4 Practical Examples

1. Passing a URL as a query parameter

const targetUrl = 'https://example.com/search?q=hello&lang=en';
const link = `/redirect?url=${encodeURIComponent(targetUrl)}`;
// "/redirect?url=https%3A%2F%2Fexample.com%2Fsearch%3Fq%3Dhello%26lang%3Den"

2. Encoding JSON data in a URL

const filters = JSON.stringify({ category: 'A&B', price: '<100' });
const url = `/products?f=${encodeURIComponent(filters)}`;
// "/products?f=%7B%22category%22%3A%22A%26B%22%2C%22price%22%3A%22%3C100%22%7D"

3. Building a mailto link with special characters

const subject = 'Questions & Answers';
const body = 'What is 50% off?";
const mailto = `mailto:help@example.com?subject=${encodeURIComponent(subject)}&body=${encodeURIComponent(body)}`;

4. Avoiding double-encoding

// WRONG: Double-encoding produces %2520 instead of %20
const already = encodeURIComponent('hello world'); // "hello%20world"
const double = encodeURIComponent(already);         // "hello%2520world"

// CORRECT: Check before encoding
function safeEncode(str) {
  try {
    return encodeURIComponent(decodeURIComponent(str));
  } catch {
    return encodeURIComponent(str);
  }
}

Related Guides

• What Is URL Encoding
• URL Encoding vs Base64
• Base64URL Difference

Try It Yourself

Encode and decode any string instantly with our free URL Encoder/Decoder tool. Perfect for debugging URLs, testing API parameters, or learning percent-encoding hands-on.