IP Address Classes: A, B, C, D, and E Explained

IP address classes were the first method used to organize the IPv4 address space into manageable blocks. Introduced in 1981 as part of the original Internet Protocol specification, the classful system divided all IP addresses into five categories: Class A through Class E. Each class defined a fixed network size and determined how many networks and hosts were possible within that range.

The system was simple and worked when the internet had a few thousand connected networks. By the early 1990s, it was clear that classful addressing was burning through the IPv4 address supply at an unsustainable rate. CIDR replaced it in 1993, but the class terminology remains deeply embedded in networking vocabulary.

Class A: Large Networks

Class A addresses occupy the range 1.0.0.0 through 126.255.255.255. The first bit of any Class A address is always 0, which is why the range starts at 1 and ends at 126 (the range 0.x.x.x is reserved, and 127.x.x.x is reserved for loopback).

In a Class A network, the first octet identifies the network, and the remaining three octets identify individual hosts. The default subnet mask is 255.0.0.0 (/8). Each Class A network can hold up to 16,777,214 hosts.

Only 126 Class A networks exist. These were allocated to the largest organizations and government agencies in the early days of the internet. Companies like General Electric (3.0.0.0/8), AT&T (12.0.0.0/8), Apple (17.0.0.0/8), and the U.S. Department of Defense (multiple blocks) received Class A allocations. Each of these organizations controls over 16 million IP addresses.

The private Class A range is 10.0.0.0 through 10.255.255.255. This entire /8 block is reserved for private network use. Large enterprises, cloud providers, and data centres commonly use 10.x.x.x addressing for internal infrastructure because the space is large enough to accommodate millions of devices across many subnets.

Class B: Medium Networks

Class B addresses cover the range 128.0.0.0 through 191.255.255.255. The first two bits of a Class B address are always 10 (in binary), which defines the range boundaries.

In a Class B network, the first two octets identify the network, and the last two identify hosts. The default subnet mask is 255.255.0.0 (/16). Each Class B network supports up to 65,534 hosts.

There are 16,384 possible Class B networks. These were allocated to medium-sized organizations, including universities, mid-size corporations, and government agencies. Universities were among the heaviest recipients of Class B allocations during the internet’s early growth.

The private Class B range is 172.16.0.0 through 172.31.255.255 (172.16.0.0/12). This provides 1,048,576 addresses split across 16 Class B-sized blocks. Medium business networks, some cloud platforms, and Docker’s default networking use addresses from this range.

Class B was where the waste problem was most visible. An organization with 500 workstations received a block of 65,534 addresses. The remaining 65,000+ addresses sat unused but could not be reallocated. As demand for internet connectivity exploded in the late 1980s, the Class B pool depleted rapidly.

Class C: Small Networks

Class C addresses span 192.0.0.0 through 223.255.255.255. The first three bits are always 110 (in binary).

In a Class C network, the first three octets identify the network, and the last octet identifies hosts. The default subnet mask is 255.255.255.0 (/24). Each Class C network supports up to 254 hosts.

Over 2 million Class C networks exist (2,097,152 to be precise). These were allocated to small organizations that needed a modest number of addresses. The limited host count (254) made Class C networks suitable for small offices and branch locations.

The private Class C range is 192.168.0.0 through 192.168.255.255 (192.168.0.0/16). This is the range used by virtually every home router for local network addressing. Your home network’s 192.168.1.x or 192.168.0.x addresses come from this range. Despite being called “Class C” colloquially, the full 192.168.0.0/16 block actually spans 256 Class C networks.

Class D and Class E

Class D and Class E are special-purpose ranges that are not used for standard host addressing.

Class D (224.0.0.0 to 239.255.255.255) is reserved for multicast. Multicast delivers a single data stream to multiple recipients simultaneously without the sender needing to transmit individual copies to each one. IPTV, video conferencing, stock market data feeds, and routing protocols like OSPF use multicast addresses. The address 224.0.0.1 is “all hosts on this subnet” and 224.0.0.2 is “all routers on this subnet.”

Class E (240.0.0.0 to 255.255.255.255) was reserved for experimental and future use. This block was never allocated for public addressing. The address 255.255.255.255 serves as the limited broadcast address (broadcast to all devices on the local network segment). Efforts have been made to reclaim the Class E space for general use as IPv4 addresses became scarce, but the lack of consistent support across networking equipment and operating systems has limited adoption.

Neither Class D nor Class E addresses are assigned to individual devices, routed across the internet in the same way as A/B/C addresses, or relevant to home networking configuration.

Why Classful Addressing Was Replaced

The classful system collapsed under a fundamental design flaw: the gap between class sizes was too large.

A Class C network provided 254 hosts. A Class B network provided 65,534 hosts. There was nothing in between. An organization that needed 1,000 addresses had two options: request four Class C networks (administrative headache, fragmented routing) or request one Class B network (massive waste). Most chose Class B.

This created two cascading problems. First, the Class B pool was being exhausted far faster than expected. Second, the global routing table was growing uncontrollably as organizations with multiple Class C allocations each required separate routing entries.

CIDR solved both problems by introducing variable-length subnet masks. With CIDR, that organization needing 1,000 addresses could receive a /22 block (1,022 usable addresses) instead of wasting a /16. Addresses were allocated precisely to match actual need. CIDR also enabled route aggregation, allowing ISPs to advertise one summary route instead of hundreds of individual Class C routes.

Today, classful addressing is a historical concept taught in networking courses and referenced on certification exams. Modern networking uses CIDR exclusively. But the terminology survives: people still say “Class C” when they mean a /24 subnet, and the private address ranges are still described by their original class designations.

Private Ranges Per Class: Summary

The three private IP ranges defined by RFC 1918 map directly to the old class system:

Class	Private Range	CIDR	Addresses	Typical Use
A	10.0.0.0 - 10.255.255.255	10.0.0.0/8	16,777,216	Large enterprise, cloud, data centre
B	172.16.0.0 - 172.31.255.255	172.16.0.0/12	1,048,576	Medium business, container networks
C	192.168.0.0 - 192.168.255.255	192.168.0.0/16	65,536	Home networks, small offices

These ranges remain in universal daily use. Every home router, every corporate network, and every cloud virtual network relies on RFC 1918 private addressing. The classes may be gone, but their most practical legacy lives on in every network you connect to.