A proposal, set forth in RFC 1519, to allocate IP addresses so as to allow the addresses to be aggregated when advertised as routes. It is based on the elimination of intrinsic IP network addresses; that is, the determination of the network address based on the first few bits of the IP address. See also: IP address, network address, supernet. [Source: RFC1983

A scheme for dividing the network and host portions of an IP v4 address on a bit-by-bit basis. Replaces the original byte-delimited class A,B,C,D IP address method, and provides greater flexibility in dividing a finite number of IP addresses among many subnets.

Historically, Internet Protocol (IP) network numbers have been divided into classes, which indicate how many individual hosts, or machines, the network number can support. For example, a Class A IP network number can support 16,777,214 hosts; a Class B IP network number can support 65,532 hosts; and a Class C IP network number can support 254 hosts. Classless Inter-Domain Routing is an addressing and routing scheme that uses a group of contiguous Class C addresses in place of a Class B address. CIDR allows for the aggregation of address space so that routing table size can be kept to a minimum, and also enables the hardware and software that route data between different portions of the Internet to more easily deal with the task. CIDR helps to enable the allocation of IP network numbers in a manner that supports the growing community of network users.