HTML Namespaces and Namespace Declarations (xmlns=)


ATTENTION: THIS PAGE IS Valid HTML 5 AND IS BEST VIEWED WITH HTML 5 - Please upgrade your browser or download one of the HTML 5 compatible browsers such as Mozilla Firefox, Chrome, Opera or IE 9 (March 14, 2011 or later). For more information see HTML 5 browsers.

If you find this helpful, please click the Google +1 Button to the left, if it is white, to make it turn blue or red. Thank you! (It also helps find this page again more easily.)

PDF mobile

HTML Namespace Definitions

Definition: HTML Namespace(s)

(Some links on this page take you to details in the HTML Tag Reference. Bookmark this page in your Favorites so you can come back to it later.)

A namespace is a property associated with the name of some object, such as a node in an HTML document, that distinguishes the type of object represented by the name from other types of objects with the same name in other namespaces. For example, in

<head><title>One Of My Web Pages</title></head>

title is the name of an element (the title element) in HTML while in

<a href="..." title="One of my links" ...>

title is the name of a title attribute on an HTML <a> tag.

The HTML namespaces (plural) in general are the collection of various namespaces in HTML code. The HTML namespace (singular) itself is the one associated with the namespace URI

On this site, the notation <a href> refers to the href attribute in the html:a element namespace, where html: is the namespace prefix for the HTML namespace.

Why are there namespaces in HTML?

Names that are spelled the same and have the same namespace cannot be distinguished from one another without knowing the context in which they appear. For example, in the RSS version 2.0 and OPML version 2.0 code below, the three <title> elements with the same name in the same namespace partition (the element partition without a namespace URI in this case) all identify the same type of object:

<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0">
      <title>Feed Title</title>
         <title>Item Title</title>

<?xml version="1.0" encoding="UTF-8"?>
<opml version="1.0">
      <title>My OPML</title>

In order to differentiate the title elements above you would have to know that the first one was a descendant of an rss element and channel element; the second was a descendant of an rss element, channel element and item element; and the third was a descendant of an opml element and a head element.

In addition, without namespaces there might be elements with the same element name that have different meanings or content depending on where they are used. Their different models would have to be combined to create a single definition (XML Schema definition or DTD) of the element. This might work if all meanings were determined by the same entity, but it would likely be a major issue if different standards bodies or industry groups were involved.

This type of confusion in XML content without namespaces makes it very difficult to aggregate that content into other types of content, even if the other content formats do support namespaces. For example, if portions of the content above was included in an HTML document (without overriding the namespace with xmlns=""), the title elements would be treated as a valid HTML titles, the head element and body element would be treated as duplicates of their HTML counterparts and most of the other elements would be treated as invalid HTML code.

Note that the HTML title element would not be confused with the title elements shown above, since its name is in a different namespace partition associated with the HTML namespace URI.

Namespace Partitions

For each namespace, including the one with no namespace URI, there are a number of namespace partitions:

  • an element name partition for the names of element types
  • for each element type, a separate element attribute name partition for the names of local attributes specific to that element type
  • a global attribute name partition for the names of global attributes that can be used on any type of element, regardless of which namespace contains the element name

Since those partitions exist for each namespace, the collection of namespace partitions can be viewed as a two-dimensional table:

Namespace URI
(in xmlns attribute)
PrefixElement NamesElement Attribute NamesGlobal Attribute Names
none (xmlns="")nonetitle (in RSS, OPML, etc.)<rss version>
<opml version>
HTML global attributes such as id and style (no prefix)<a> <link> <form><a href> <a rel>
<link href> <link rel>
<form action> <form method>
default namespace (xmlns=) and namespace prefix declarations (xmlns:xxx=) for the HTML examples on this site<element> <attribute><element name> <element ref>
<attribute name> <attribute ref> for the HTML examples on this sitexsi:type for the HTML examples on this site<math> <mo><math mode>
<mo form> for the HTML examples on this siterdf:RDF rdf:Description rdf:Listrdf:about for the HTML examples on this siterdfs:Datatype for the HTML examples on this site<svg> <rect><svg viewBox>
<rect width> <rect height>

When designing schema for one or more sets of objects, you should ensure that no namespace partition (as shown in the table above) contains the same name for two different object types. Otherwise, it will be difficult to create a definition (XML Schema definition or DTD) that properly describes the schema. For example, in the table above the title element in the "no namespace URL" partition violates this requirement because it is used for RSS channel titles, RSS item titles and OPMN titles.
and other HTML Namespaces

The namespace URIs in HTML documents

Commonly used namespaces in HTML include:
the URI for the XML base namespace; associated with the implicitly declared reserved prefix xml:
the namespace URI for XML namespace declarations; associated with the implicitly declared reserved prefix xmlns:
the URI for the HTML namespace, the same namespace URI defined by the 2000-2010 Recommendations from the W3C HTML Working Group; implicitly declared as the default namespace for unprefixed tag names when document is being parsed as HTML, but explicitly coding it is recommended for cases when document is being parsed as xHTML or XML
the MathML namespace for MathML tags
the RDF namespace for RDF tags
the RDF schema namespace
the SVG namespace for SVG tags
the XLink namespace
the namespace for XML Schema Definitions, which are XML documents that replace non-XML DTDs
the namespace for XML Schema instance documents, which can be used to specify whether the data for a field is binary (possibly encrypted) or plain text:
<span id="masked-credit-card-number" xsi:type="xsd:string">4321 **** **** 8765</span>
<span id="encrypted-credit-card-number" xsi:type="xsd:base64Binary">BAM0NComFzC2TOsmRzW0NTueQU==</span>
the namespace for XSL style sheets, which is not used in HTML documents themselves, but is used in the style sheet documents that are referenced by the xml-stylesheet instruction and contain XSLT elements intermingled with HTML code.

xmlns namespace declarations in the <html> tag

How to declare namespaces in an HTML document

The easiest way to declare namespaces is by putting the xmlns attributes in the top element of the XML document, which in this case is the <html> tag:

<html xmlns=""

In HTML 5, all elements (HTML tags) are automatically considered to be qualified with the HTML 5 namespace, which makes the declaration of the HTML namespace optional, but this only works with HTML browsers with HTML 5 support that are parsing the web page as an HTML version 5 document. It's best to continue coding the xmlns="" xmlns attribute explicitly to provide backward compatibility with browsers lacking full HTML 5 support and other types of programs that may be processing HTML documents, such as RSS feed readers and generic XML parsers - otherwise all of the HTML tags will appear to be in the partition with no namespace URI.

Valid HTML 5