Documentation for this section has not yet been entered.
The purpose of this property is to enable encoding of the carriage return and newline characters, \r and \n, that are found in response headers.
In outbound response headers, the characters that are represented by the codes 0x1F and below are encoded and also the character 0x7F (delete character). The only exception is that the character 0x09 (the tab character) is unmodified.
This encoding can help to avoid injection attacks that exploit an application that echoes untrusted data contained by the header.
This property does not apply to the status line itself (status code and status description), but should apply to other headers. Although <httpRuntime> can be set at any level, this property is only applicable at the machine and application level.
When this property is true, which is the default, the \r or \n characters found in a response header are encoded to %0d and %0a. This defeats header-injection attacks by making the injected material part of the same header line. This might break the response but should not open attack vectors against the client. Echoing back untrusted data is never a good idea in any situation, though.
HTTP header continuations rely on headers spanning multiple lines and require new lines in them. If you need to use header continuations, you need to set the HttpRuntimeSection.EnableHeaderChecking property to false. Because there is a performance impact from looking at headers, if you are certain you are already doing the right checks, turning off this feature can improve the performance of your application. Before you disable this feature, be sure you are already taking the right precautions in this area.