Why does an IP address beginning with 10 belong to Class A, while one beginning with 192 belongs to Class C?
The answer comes from an early system for organizing IPv4 addresses. This system divided the address space into five classes: A, B, C, D and E.
Understanding these classes helps you read networking documentation and learn how network addresses work. However, modern networks use explicit subnet masks and prefix lengths, so an address class alone does not tell you the actual size of a subnet.
Let’s explore the five classes, their ranges, and what they mean today.
What Is an IPv4 Address?
An IPv4 address contains 32 bits, usually written as four decimal numbers separated by dots:
192.168.1.10
Each number represents an octet, or eight bits. An octet can have a decimal value from 0 to 255.
IPv4 therefore has approximately 4.3 billion possible address combinations. Some are reserved for special purposes and cannot be assigned as ordinary device addresses.
For a typical unicast address, the subnet mask or prefix length determines which bits identify the network and which identify a host within it.
The original classful system used fixed boundaries for three different network sizes. These address structures were defined in the original IPv4 specification, RFC 791.
IPv4 Address Classes at a Glance
The historical class of an IPv4 address is determined by its first octet.
| Class | First-octet range | Leading binary bits | Historical default mask | Main purpose |
|---|---|---|---|---|
| A | 0–127* | 0 |
255.0.0.0, or /8 | Large unicast networks |
| B | 128–191 | 10 |
255.255.0.0, or /16 | Medium-sized unicast networks |
| C | 192–223 | 110 |
255.255.255.0, or /24 | Small unicast networks |
| D | 224–239 | 1110 |
Not applicable | Multicast |
| E | 240–255 | 1111 |
Not applicable | Reserved address space |
The Class A bit pattern covers first-octet values from 0 to 127, but 0 and 127 have special uses. Ordinary Class A network numbers are traditionally listed as 1–126.
The binary prefixes are read as individual bits. For example, Class B begins with “one, zero,” rather than the decimal number ten.
The fixed leading bits create the boundaries between the classes.
Class A: Large Networks
Class A addresses begin with a binary 0. Their historical default prefix is /8, equivalent to the subnet mask:
255.0.0.0
Under this default mask, the first octet forms the network portion, and the remaining three octets form the host portion.
Consider:
10.1.2.3/8
Its structure is:
- Network portion:
10 - Host portion:
1.2.3 - Network address:
10.0.0.0
A default-sized Class A network contains 24 host bits, providing 16,777,216 total addresses. Under conventional network and broadcast address rules, 16,777,214 are usable host addresses.
This explains why Class A was intended for very large networks.
Two special ranges deserve attention. Addresses beginning with 127 are used for loopback, allowing a device to communicate with itself. The 0.0.0.0/8 block also has special meanings and is not an ordinary assignable network. These reservations are documented in the IANA IPv4 Special-Purpose Address Registry.
Class B: Medium-Sized Networks
Class B addresses have first-octet values from 128 through 191. Their leading binary bits are 10.
The historical default prefix is /16, with this mask:
255.255.0.0
The first two octets form the network portion, leaving two octets for hosts.
For example:
172.16.1.10/16
Under that prefix:
- Network portion:
172.16 - Host portion:
1.10 - Network address:
172.16.0.0
Sixteen host bits provide 65,536 total addresses, or 65,534 conventionally usable host addresses.
Class B offered a size between the very large Class A networks and the smaller Class C networks.
Class C: Small Networks
Class C addresses have first-octet values from 192 through 223. Their leading binary bits are 110.
The historical default prefix is /24, equivalent to:
255.255.255.0
The first three octets form the network portion, and the final octet identifies the host.
Consider:
192.168.1.10/24
Its network address is 192.168.1.0.
A conventional /24 subnet has 256 total addresses:
192.168.1.0identifies the network.192.168.1.1through192.168.1.254are usable host addresses.192.168.1.255is the subnet broadcast address.
That leaves 254 usable host addresses.
These calculations apply to the stated prefix. A Class C first octet does not force an address to use /24 in a modern network.
Class D: Multicast
Class D addresses have first-octet values from 224 through 239, with leading binary bits of 1110.
They identify multicast groups, rather than ordinary individual hosts.
For example:
239.1.1.1
A multicast destination allows a sender to address a group of receivers. Network devices and receiving hosts must support and be configured for the relevant multicast behavior.
Class D does not use the historical default network-and-host masks associated with Classes A, B and C. Its group-address structure is defined in RFC 1112.
Class E: Reserved Addresses
Class E addresses have first-octet values from 240 through 255, beginning with binary 1111.
They are often described in introductory material as experimental. Reserved address space is a more accurate general description.
These addresses are not generally available for ordinary public host assignment.
One specific address requires separate treatment:
255.255.255.255
This is the IPv4 limited broadcast address. It has a defined broadcast purpose, even though its first octet falls within the historical Class E range. Both the reserved block and limited broadcast address appear in the IANA special-purpose registry.
How to Identify an Address Class
For historical classification, check only the first octet.
| Address | First octet | Historical class |
|---|---|---|
| 10.1.2.3 | 10 | A |
| 172.16.1.1 | 172 | B |
| 192.168.1.1 | 192 | C |
| 239.1.1.1 | 239 | D |
| 250.1.1.1 | 250 | E |
Recognizing the class does not establish whether an address is public, private, reserved, or suitable for a particular device. Those are separate questions.
Address Classes vs. Modern CIDR
Modern networks use Classless Inter-Domain Routing, abbreviated CIDR.
CIDR makes the prefix length explicit, allowing networks to use sizes beyond the historical class boundaries. The approach is described in RFC 4632.
Compare these two addresses:
| Address and prefix | Network address | Host bits |
|---|---|---|
| 10.1.2.3/8 | 10.0.0.0 | 24 |
| 10.1.2.3/24 | 10.1.2.0 | 8 |
Both addresses have a historical Class A first octet. However, their configured prefixes create different network boundaries.
When configuring a server, troubleshooting connectivity, or calculating a subnet, use the supplied prefix or subnet mask.
Common Questions
Are all Class A addresses private?
No. Address class and private addressing describe different properties. An address beginning with 10 belongs to a private range, but other Class A ranges have different uses.
Does Class C always mean a /24 subnet?
No. /24 is its historical default. Modern configurations can specify another prefix length.
Can a Class D address be assigned as a normal server address?
A Class D address identifies a multicast group. It is not an ordinary unicast server address.
Do address classes apply to IPv6?
No. Classes A through E belong to IPv4. IPv6 uses a different address format and addressing architecture.
What to Remember
The first octet tells you an IPv4 address’s historical class. Classes A, B and C used default prefixes of /8, /16 and /24. Class D identifies multicast groups, while Class E covers reserved space.
For practical networking work, always check the actual prefix length. It determines the subnet your device or server belongs to.
