他支持以下DOM2方法:
appendChild, cloneNode, hasAttributes, hasChildNodes, insertBefore, normalize, removeChild, replaceChild.
也支持以下DOM2屬性:
attributes, childNodes, firstChild, lastChild, localName, namespaceURI, nextSibling, nodeName, nodeType, nodeValue, ownerDocument, parentNode, prefix, previousSibling, textContent.
其他方法可以將documentFragment 作為一個參數，（比如Node的 appendChild和insertBefore 方法），這樣，fragment 就可以被追加到父對象中。
Example：
复制代码 代码如下:
var frag = document.createDocumentFragment();
frag.appendChild(document.createTextNode('Ipsum Lorem'));
document.body.appendChild(frag);

document.createDocumentFragment()说白了就是为了节约使用DOM。每次JavaScript对DOM的操作都会改变页面的变现，并重新刷新整个页面，从而消耗了大量的时间。为解决这个问题，可以创建一个文档碎片，把所有的新节点附加其上，然后把文档碎片的内容一次性添加到document中。
复制代码 代码如下:
var oui=document.getElementById("oItem");
for(var i=0;i<10;i++)
{
var oli=document.createElement("li");
oui.appendChild(oli);
oli.appendChild(document.createTextNode("Item"+i));
}

上面的代码在循环中调用了oui.appendChild（oli），每次执行这条语句后，浏览器都会更新页面。其次下面的oui.appendChild()添加了文本节点，也要更新页面。所以一共要更新页面20次。
为了页面的优化，我们要尽量减少DOM的操作，将列表项目在添加文本节点之后再添加，并合理地使用creatDocumentFragment(),代码如下：
复制代码 代码如下:
var oui=document.getElementById("oItem");
var oFragment=document.createDocumentFragment();
for(var i=0;i<10;i++){
var oli=document.createElement("li");
oli.appendChild(document.createTextNode("Item"+i));
oFragment.appendChild(oli);
}
oui.appendChild(oFragment);

W3C参考：http://mand like cut or rearranging a document by moving fragments around. It is desirable to have an object which can hold such fragments and it is quite natural to use a Node for this purpose. While it is true that a Document object could fulfill this role, a Document object can potentially be a heavyweight object, depending on the underlying implementation. What is really needed for this is a very lightweight object. DocumentFragment is such an object.

Furthermore, various operations -- such as inserting nodes as children of another Node -- may take DocumentFragment objects as arguments; this results in all the child nodes of the DocumentFragment being moved to the child list of this node.

The children of a DocumentFragment node are zero or more nodes representing the tops of any sub-trees defining the structure of the document. DocumentFragment nodes do not need to be well-formed XML documents (although they do need to follow the rules imposed upon well-formed XML parsed entities, which can have multiple top nodes). For example, a DocumentFragment might have only one child and that child node could be a Text node. Such a structure model represents neither an HTML document nor a well-formed XML document.

When a DocumentFragment is inserted into a Document (or indeed any other Node that may take children) the children of the DocumentFragment and not the DocumentFragment itself are inserted into the Node. This makes the DocumentFragment very useful when the user wishes to create nodes that are siblings; the DocumentFragment acts as the parent of these nodes so that the user can use the standard methods from the Node interface, such as insertBefore and appendChild.