CSS is designed to enable the separation of presentation and content, including layout, colors, and fonts. This separation can improve content accessibility, provide more flexibility and control in the specification of presentation characteristics, enable multiple web pages to share formatting by specifying the relevant CSS in a separate .css file which reduces complexity and repetition in the structural content as well as enabling the .css file to be cached to improve the page load speed between the pages that share the file and its formatting.
Separation of formatting and content also makes it feasible to present the same markup page in different styles for different rendering methods, such as on-screen, in print, by voice (via speech-based browser or screen reader), and on Braille-based tactile devices. CSS also has rules for alternate formatting if the content is accessed on a mobile device.
The name cascading comes from the specified priority scheme to determine which style rule applies if more than one rule matches a particular element. This cascading priority scheme is predictable.
The CSS specifications are maintained by the World Wide Web Consortium (W3C). Internet media type (MIME type)
text/css is registered for use with CSS by RFC 2318 (March 1998). The W3C operates a free CSS validation service for CSS documents.
CSS has a simple syntax and uses a number of English keywords to specify the names of various style properties.
In CSS, selectors declare which part of the markup a style applies to by matching tags and attributes in the markup itself.
Classes and IDs are case-sensitive, start with letters, and can include alphanumeric characters, hyphens, and underscores. A class may apply to any number of instances of any elements. An ID may only be applied to a single element.
Pseudo-classes are used in CSS selectors to permit formatting based on information that is not contained in the document tree. One example of a widely used pseudo-class is
:hover, which identifies content only when the user "points to" the visible element, usually by holding the mouse cursor over it. It is appended to a selector as in
#elementid:hover. A pseudo-class classifies document elements, such as
:visited, whereas a pseudo-element makes a selection that may consist of partial elements, such as
The following table provides a summary of selector syntax indicating usage and the version of CSS that introduced it.
A declaration block consists of a list of declarations in braces. Each declaration itself consists of a property, a colon (
:), and a value. If there are multiple declarations in a block, a semi-colon (
;) must be inserted to separate each declaration. An optional semi-colon after the last (or single) declaration may be used.
Properties are specified in the CSS standard. Each property has a set of possible values. Some properties can affect any type of element, and others apply only to particular groups of elements.
Values may be keywords, such as "center" or "inherit", or numerical values, such as
200px (200 pixels),
50vw (50 percent of the viewport width) or 80% (80 percent of the parent element's width). Color values can be specified with keywords (e.g. "
red"), hexadecimal values (e.g.
#FF0000, also abbreviated as
#F00), RGB values on a 0 to 255 scale (e.g.
), RGBA values that specify both color and alpha transparency (e.g.
rgb(255, 0, 0)
rgba(255, 0, 0, 0.8)), or HSL or HSLA values (e.g.
hsl(000, 100%, 50%),
hsla(000, 100%, 50%, 80%)).
Non-zero numeric values representing linear measures must include a length unit, which is either an alphabetic code or abbreviation, as in
50vw; or a percentage sign, as in
80%. Some units –
pc (pica); and
pt (point) – are absolute, which means that the rendered dimension does not depend upon the structure of the page; others –
ex (ex) and
px (pixel) – are relative, which means that factors such as the font size of a parent element can affect the rendered measurement. These eight units were a feature of CSS 1 and retained in all subsequent revisions. The proposed CSS Values and Units Module Level 3 will, if adopted as a W3C Recommendation, provide seven further length units:
Before CSS, nearly all presentational attributes of HTML documents were contained within the HTML markup. All font colors, background styles, element alignments, borders and sizes had to be explicitly described, often repeatedly, within the HTML. CSS lets authors move much of that information to another file, the style sheet, resulting in considerably simpler HTML.
For example, headings (
h1 elements), sub-headings (
h2), sub-sub-headings (
h3), etc., are defined structurally using HTML. In print and on the screen, choice of font, size, color and emphasis for these elements is presentational.
Before CSS, document authors who wanted to assign such typographic characteristics to, say, all
h2 headings had to repeat HTML presentational markup for each occurrence of that heading type. This made documents more complex, larger, and more error-prone and difficult to maintain. CSS allows the separation of presentation from structure. CSS can define color, font, text alignment, size, borders, spacing, layout and many other typographic characteristics, and can do so independently for on-screen and printed views. CSS also defines non-visual styles, such as reading speed and emphasis for aural text readers. The W3C has now deprecated the use of all presentational HTML markup.
For example, under pre-CSS HTML, a heading element defined with red text would be written as:
Using CSS, the same element can be coded using style properties instead of HTML presentational attributes:
The advantages of this may not be immediately clear but the power of CSS becomes more apparent when the style properties are placed in an internal style element or, even better, an external CSS file. For example, suppose the document contains the style element:
h1 elements in the document will then automatically become red without requiring any explicit code. If the author later wanted to make
h1 elements blue instead, this could be done by changing the style element to:
rather than by laboriously going through the document and changing the color for each individual
The styles can also be placed in an external CSS file, as described below, and loaded using syntax similar to:
This further decouples the styling from the HTML document and makes it possible to restyle multiple documents by simply editing a shared external CSS file.
CSS information can be provided from various sources. These sources can be the web browser, the user, and the author. The information from the author can be further classified into inline, media type, importance, selector specificity, rule order, inheritance, and property definition. CSS style information can be in a separate document, or it can be embedded into an HTML document. Multiple style sheets can be imported. Different styles can be applied depending on the output device being used; for example, the screen version can be quite different from the printed version, so that authors can tailor the presentation appropriately for each medium.
The style sheet with the highest priority controls the content display. Declarations not set in the highest priority source are passed on to a source of lower priority, such as the user agent style. The process is called cascading.
One of the goals of CSS is to allow users greater control over presentation. Someone who finds red italic headings difficult to read may apply a different style sheet. Depending on the browser and the web site, a user may choose from various style sheets provided by the designers, or may remove all added styles and view the site using the browser's default styling, or may override just the red italic heading style without altering other attributes.
Specificity refers to the relative weights of various rules. It determines which styles apply to an element when more than one rule could apply. Based on specification, a simple selector (e.g. H1) has a specificity of 1, class selectors have a specificity of 1,0, and ID selectors a specificity of 1,0,0. Because the specificity values do not carry over as in the decimal system, commas are used to separate the "digits" (a CSS rule having 11 elements and 11 classes would have a specificity of 11,11, not 121).
Thus the following rules selectors result in the indicated specificity:
In the above example, the declaration in the
style attribute overrides the one in the
<style> element because it has a higher specificity, and thus, the paragraph appears green.
Inheritance is a key feature in CSS; it relies on the ancestor-descendant relationship to operate. Inheritance is the mechanism by which properties are applied not only to a specified element, but also to its descendants. Inheritance relies on the document tree, which is the hierarchy of XHTML elements in a page based on nesting. Descendant elements may inherit CSS property values from any ancestor element enclosing them. In general, descendant elements inherit text-related properties, but their box-related properties are not inherited. Properties that can be inherited are color, font, letter-spacing, line-height, list-style, text-align, text-indent, text-transform, visibility, white-space and word-spacing. Properties that cannot be inherited are background, border, display, float and clear, height, and width, margin, min- and max-height and -width, outline, overflow, padding, position, text-decoration, vertical-align and z-index.
Inheritance can be used to avoid declaring certain properties over and over again in a style sheet, allowing for shorter CSS.
Inheritance in CSS is not the same as inheritance in class-based programming languages, where it is possible to define class B as "like class A, but with modifications". With CSS, it is possible to style an element with "class A, but with modifications". However, it is not possible to define a CSS class B like that, which could then be used to style multiple elements without having to repeat the modifications.
Suppose there is an h1 element with an emphasizing element (em) inside:
If no color is assigned to the em element, the emphasized word "illustrate" inherits the color of the parent element, h1. The style sheet h1 has the color pink, hence, the em element is likewise pink.
Whitespace between properties and selectors is ignored. This code snippet:
One common way to format CSS for readability is to indent each property and give it its own line. In addition to formatting CSS for readability, shorthand properties can be used to write out the code faster, which also gets processed more quickly when being rendered:
There are five possible values of the
position property. If an item is positioned in any way other than
static, then the further properties
right are used to specify offsets and positions.The element having position static is not affected by the
float property may have one of three values. Absolutely positioned or fixed items cannot be floated. Other elements normally flow around floated items, unless they are prevented from doing so by their
CSS was first proposed by Håkon Wium Lie on October 10, 1994. At the time, Lie was working with Tim Berners-Lee at CERN. Several other style sheet languages for the web were proposed around the same time, and discussions on public mailing lists and inside World Wide Web Consortium resulted in the first W3C CSS Recommendation (CSS1) being released in 1996. In particular, a proposal by Bert Bos was influential; he became co-author of CSS1, and is regarded as co-creator of CSS.
Style sheets have existed in one form or another since the beginnings of Standard Generalized Markup Language (SGML) in the 1980s, and CSS was developed to provide style sheets for the web. One requirement for a web style sheet language was for style sheets to come from different sources on the web. Therefore, existing style sheet languages like DSSSL and FOSI were not suitable. CSS, on the other hand, let a document's style be influenced by multiple style sheets by way of "cascading" styles.
As HTML grew, it came to encompass a wider variety of stylistic capabilities to meet the demands of web developers. This evolution gave the designer more control over site appearance, at the cost of more complex HTML. Variations in web browser implementations, such as ViolaWWW and WorldWideWeb, made consistent site appearance difficult, and users had less control over how web content was displayed. The browser/editor developed by Tim Berners-Lee had style sheets that were hard-coded into the program. The style sheets could therefore not be linked to documents on the web. Robert Cailliau, also of CERN, wanted to separate the structure from the presentation so that different style sheets could describe different presentation for printing, screen-based presentations, and editors.
Improving web presentation capabilities was a topic of interest to many in the web community and nine different style sheet languages were proposed on the www-style mailing list. Of these nine proposals, two were especially influential on what became CSS: Cascading HTML Style Sheets and Stream-based Style Sheet Proposal (SSP). Two browsers served as testbeds for the initial proposals; Lie worked with Yves Lafon to implement CSS in Dave Raggett's Arena browser. Bert Bos implemented his own SSP proposal in the Argo browser. Thereafter, Lie and Bos worked together to develop the CSS standard (the 'H' was removed from the name because these style sheets could also be applied to other markup languages besides HTML).
Development of HTML, CSS, and the DOM had all been taking place in one group, the HTML Editorial Review Board (ERB). Early in 1997, the ERB was split into three working groups: HTML Working group, chaired by Dan Connolly of W3C; DOM Working group, chaired by Lauren Wood of SoftQuad; and CSS Working group, chaired by Chris Lilley of W3C.
The CSS Working Group began tackling issues that had not been addressed with CSS level 1, resulting in the creation of CSS level 2 on November 4, 1997. It was published as a W3C Recommendation on May 12, 1998. CSS level 3, which was started in 1998, is still under development as of 2014.
In 2005, the CSS Working Groups decided to enforce the requirements for standards more strictly. This meant that already published standards like CSS 2.1, CSS 3 Selectors, and CSS 3 Text were pulled back from Candidate Recommendation to Working Draft level.
The CSS 1 specification was completed in 1996. Microsoft's Internet Explorer 3 was released in that year, featuring some limited support for CSS. IE 4 and Netscape 4.x added more support, but it was typically incomplete and had many bugs that prevented CSS from being usefully adopted. It was more than three years before any web browser achieved near-full implementation of the specification. Internet Explorer 5.0 for the Macintosh, shipped in March 2000, was the first browser to have full (better than 99 percent) CSS 1 support, surpassing Opera, which had been the leader since its introduction of CSS support fifteen months earlier. Other browsers followed soon afterward, and many of them additionally implemented parts of CSS 2.
However, even when later 'version 5' web browsers began to offer a fairly full implementation of CSS, they were still incorrect in certain areas and were fraught with inconsistencies, bugs and other quirks. Microsoft Internet Explorer 5.x for Windows, as opposed to the very different IE for Macintosh, had a flawed implementation of the 'CSS box model', as compared with the CSS standards. Such inconsistencies and variation in feature support made it difficult for designers to achieve a consistent appearance across browsers and platforms without the use of workarounds termed CSS hacks and filters. The IE/Windows box model bugs were so serious that, when Internet Explorer 6 was released, Microsoft introduced a backwards-compatible mode of CSS interpretation ('quirks mode') alongside an alternative, corrected 'standards mode'. Other non-Microsoft browsers also provided such 'mode'-switch behavior capability. It therefore became necessary for authors of HTML files to ensure they contained special distinctive 'standards-compliant CSS intended' marker to show that the authors intended CSS to be interpreted correctly, in compliance with standards, as opposed to being intended for the now long-obsolete IE5/Windows browser. Without this marker, web browsers that have the 'quirks mode'-switching capability will size objects in web pages as IE5/Windows would rather than following CSS standards.
Problems with patchy adoption of CSS, along with errata in the original specification, led the W3C to revise the CSS 2 standard into CSS 2.1, which moved nearer to a working snapshot of current CSS support in HTML browsers. Some CSS 2 properties that no browser successfully implemented were dropped, and in a few cases, defined behaviors were changed to bring the standard into line with the predominant existing implementations. CSS 2.1 became a Candidate Recommendation on February 25, 2004, but CSS 2.1 was pulled back to Working Draft status on June 13, 2005, and only returned to Candidate Recommendation status on July 19, 2007.
In addition to these problems, the
.css extension was used by a software product used to convert PowerPoint files into Compact Slide Show files,
so some web servers served all
.css as MIME type
application/x-pointplus rather than
Individual browser vendors occasionally introduced new parameters ahead of standardization and universalization. To prevent interfering with future implementations, vendors prepended unique names to the parameters, such as
-moz- for Mozilla Firefox,
-webkit- named after the browsing engine of Apple Safari,
-o- for Opera Browser and
-ms- for Microsoft Internet Explorer.
Occasionally, the parameters with vendor prefix such as
-webkit-linear-gradient have slightly different syntax as compared to their non-vendor-prefix counterparts.
Prefixed properties are rendered obsolete by the time of standardization. Programs are available to automatically add prefixes for older browsers, and to point out standardized versions of prefixed parameters. Since prefixes are limited to a small subset of browsers, removing the prefix allows other browsers to see the functionality. An exception is certain obsolete
-webkit- prefixed properties, which are so common and persistent on the web that other families of browsers have decided to support them for compatibility.
CSS has various levels and profiles. Each level of CSS builds upon the last, typically adding new features and typically denoted as CSS 1, CSS 2, CSS 3, and CSS 4. Profiles are typically a subset of one or more levels of CSS built for a particular device or user interface. Currently there are profiles for mobile devices, printers, and television sets. Profiles should not be confused with media types, which were added in CSS 2.
The first CSS specification to become an official W3C Recommendation is CSS level 1, published on December 17, 1996. Håkon Wium Lie and Bert Bos are credited as the original developers. Among its capabilities are support for
CSS level 2 specification was developed by the W3C and published as a recommendation in May 1998. A superset of CSS 1, CSS 2 includes a number of new capabilities like absolute, relative, and fixed positioning of elements and z-index, the concept of media types, support for aural style sheets (which were later replaced by the CSS 3 speech modules) and bidirectional text, and new font properties such as shadows.
CSS level 2 revision 1, often referred to as "CSS 2.1", fixes errors in CSS 2, removes poorly supported or not fully interoperable features and adds already implemented browser extensions to the specification. To comply with the W3C Process for standardizing technical specifications, CSS 2.1 went back and forth between Working Draft status and Candidate Recommendation status for many years. CSS 2.1 first became a on February 25, 2004, but it was reverted to a Working Draft on June 13, 2005 for further review. It returned to Candidate Recommendation on 19 July 2007 and then updated twice in 2009. However, because changes and clarifications were made, it again went back to Last Call Working Draft on 7 December 2010.
CSS 2.1 was planned as the first and final revision of level 2—but low priority work on CSS 2.2 began in 2015.
Unlike CSS 2, which is a large single specification defining various features, CSS 3 is divided into several separate documents called "modules". Each module adds new capabilities or extends features defined in CSS 2, preserving backward compatibility. Work on CSS level 3 started around the time of publication of the original CSS 2 recommendation. The earliest CSS 3 drafts were published in June 1999.
Due to the modularization, different modules have different stability and statuses.
There is no single, integrated CSS4 specification, because the specification has been split into many separate modules which level independently.
Modules that build on things from CSS Level 2 started at Level 3. Some of them have already reached Level 4 or are already approaching Level 5. Other modules that define entirely new functionality, such as Flexbox, have been designated as Level 1 and some of them are approaching Level 2.
The CSS Working Group sometimes publishes "Snapshots", a collection of whole modules and parts of other drafts that are considered stable enough to be implemented by browser developers. So far, five such "best current practices" documents have been published as Notes, in 2007, 2010, 2015, 2017, and 2018.
Since these specification snapshots are primarily intended for developers, there has been growing demand for as similar versioned reference document targeted at authors, which would present the state of interoperable implementations as meanwhile documented by sites like Can I Use… and the Mozilla Developer Network. A W3C Community Group has been established in early 2020 in order to discuss and define such a resource. The actual kind of versioning is also up to debate, which means that the document once produced might not be called "CSS4".
Each web browser uses a layout engine to render web pages, and support for CSS functionality is not consistent between them. Because browsers do not parse CSS perfectly, multiple coding techniques have been developed to target specific browsers with workarounds (commonly known as CSS hacks or CSS filters). Adoption of new functionality in CSS can be hindered by lack of support in major browsers. For example, Internet Explorer was slow to add support for many CSS 3 features, which slowed adoption of those features and damaged the browser's reputation among developers. In order to ensure a consistent experience for their users, web developers often test their sites across multiple operating systems, browsers, and browser versions, increasing development time and complexity. Tools such as BrowserStack have been built to reduce the complexity of maintaining these environments.
In addition to these testing tools, many sites maintain lists of browser support for specific CSS properties, including and the Mozilla Developer Network. Additionally, the CSS 3 defines feature queries, which provide an
Additionally, several more issues were present in prior versions of the CSS standard, but have been alleviated:
CSS frameworks are pre-prepared libraries that are meant to allow for easier, more standards-compliant styling of web pages using the Cascading Style Sheets language. CSS frameworks include Blueprint, Bootstrap, Foundation and Materialize. Like programming and scripting language libraries, CSS frameworks are usually incorporated as external .css sheets referenced in the HTML
<head>. They provide a number of ready-made options for designing and laying out the web page. Although many of these frameworks have been published, some authors use them mostly for rapid prototyping, or for learning from, and prefer to 'handcraft' CSS that is appropriate to each published site without the design, maintenance and download overhead of having many unused features in the site's styling.
As the size of CSS resources used in a project increases, a development team often needs to decide on a common design methodology to keep them organized. The goals are ease of development, ease of collaboration during development and performance of the deployed stylesheets in the browser. Popular methodologies include OOCSS (object oriented CSS), ACSS (atomic CSS), oCSS (organic Cascade Style Sheet), SMACSS (scalable and modular architecture for CSS), and BEM (block, element, modifier).