Classful IP Addressing Explained: A Guide
Learn how classful IPv4 addressing splits addresses into classes A to E, what each class is for, and why fixed blocks gave way to CIDR and subnetting.
Every device on your home network has an IP address, but few people stop to ask why the numbers look the way they do. I have seen the same confusion in every networking course: 10.0.0.1 here, 192.168.1.1 there, and no clear reason for either. The answer sits in the oldest idea in internet addressing — the class system.
This guide explains classful IP addressing, the five classes, private and public addresses, subnetting, and why the internet moved on to CIDR.
Classful addressing splits IPv4 addresses into five classes, A to E, by their first bits. Class A fits huge networks, Class B fits medium ones, Class C fits small ones, Class D is for multicast, and Class E is reserved — and your home router at 192.168.1.1 sits in a private Class C range.
What is classful addressing?
Classful addressing is the original way of organising IPv4 addresses. It divides the whole address space into five fixed classes, A to E, and each class decides how much of the address names the network and how much names the device on it.
An IPv4 address is 32 bits long, written as four numbers separated by dots, like 192.168.1.1. I read the first number to find the class, and the class tells me the size of the network at once.

The system was simple, and it worked well when the internet was small. Its fixed sizes wasted address space as networks grew, which is why it was replaced by classless addressing — but the classes still shape how we talk about networks today.
IPv4 addresses: the five classes at a glance
Each class covers a range of first numbers, a fixed split between network and host, and a set number of hosts. I keep this table beside me when I read an address, because it answers the size question in seconds.
| Class | First bits | Range | Default mask | Hosts per network | Used for |
|---|---|---|---|---|---|
| A | 0 | 1.0.0.0 – 126.255.255.255 | 255.0.0.0 | 16,777,214 | Very large networks, ISPs |
| B | 10 | 128.0.0.0 – 191.255.255.255 | 255.255.0.0 | 65,534 | Medium networks, universities |
| C | 110 | 192.0.0.0 – 223.255.255.255 | 255.255.255.0 | 254 | Small networks, homes, offices |
| D | 1110 | 224.0.0.0 – 239.255.255.255 | — | — | Multicast groups |
| E | 1111 | 240.0.0.0 – 255.255.255.255 | — | — | Reserved, experimental |
The host counts leave out the network address and the broadcast address, which is why Class C holds 254 usable hosts instead of 256. I count the same way every time, so the numbers stay honest.
Class A: definition and use cases
Class A addresses suit very large networks with many devices. The first bit is always 0, the first octet runs from 1 to 126, and each network holds over 16 million host addresses.
Very large companies and ISPs used Class A blocks, because one block could cover a whole organisation. The address 127.0.0.0 belongs to loopback inside this space, so it never counts as a normal network. I still meet Class A blocks on the big ISP networks I work with.
Class B: definition and use cases
Class B addresses suit medium-sized networks. The first two bits are 10, the first octet runs from 128 to 191, and each network holds up to 65,534 hosts.
Universities, government bodies and mid-sized organisations used Class B, because 65,000 hosts is roomy without being wasteful. A Class B network was the natural size for a campus or a city office. I still meet Class B blocks on older university and office networks.
Class C: definition and use cases
Class C addresses suit small networks such as homes and small businesses. The first three bits are 110, the first octet runs from 192 to 223, and each network holds up to 254 hosts.
Your home network is almost certainly a Class C network. The router hands out addresses like 192.168.1.2, and 254 devices is more than any home needs. I check the third octet first on home networks, because that is where Class C shows up.
Class D: multicast
Class D addresses are not for normal devices at all. They name multicast groups, where one stream goes to many receivers at once, and the first octet runs from 224 to 239.
Video streaming and conferencing use multicast, because sending one copy to the group beats sending one copy to every viewer. I never assign a Class D address to a device, because it names a group, not a host.
Class E: reserved
Class E addresses run from 240 to 255 in the first octet and are reserved for experiments. They are not for public use, and no device on a normal network should ever carry one.
Research and testing are the only jobs for Class E. If I ever see one on a live network, I treat it as a misconfiguration and fix it.
Classful networking explained
The five classes give a fixed way to split every address into a network part and a host part. Class A keeps 8 bits for the network and 24 for hosts, Class B splits 16 and 16, and Class C keeps 24 for the network and 8 for hosts.
That fixed split was the system’s strength and its flaw. It made routing simple, because a router could read the first bits and know the mask at once — but it gave a small network a whole Class C block and a big network a whole Class A block, with nothing in between. I explain the trade-off this way a lot: simple to route, wasteful to hand out.
Private vs public IP addresses
Private addresses live inside local networks and never route on the public internet, while public addresses are globally unique and let devices talk across the internet. A home router sits between the two, and NAT translates one side to the other.
| Range | Class | Where you see it |
|---|---|---|
| 10.0.0.0 – 10.255.255.255 | A | Large private networks, offices |
| 172.16.0.0 – 172.31.255.255 | B | Medium private networks |
| 192.168.0.0 – 192.168.255.255 | C | Home routers, small offices |
Private addresses are free to reuse in every home, because they never leave the local network. ISPs hand out public addresses, and there are only about 4.3 billion of them, which is why the internet ran short. I check the private range first whenever a home device will not reach the internet.
Private IPs connect the devices in your home — phones, printers, smart TVs — without exposing them to the internet. They keep the inside of your network hidden, which is a real layer of protection.
Public IPs belong to servers, websites and the router’s outside face. If you cannot reach your router at all, find your real gateway first with our Router IP Finder, because a wrong gateway address looks like a dead network.
How subnetting works
Subnetting splits a large network into smaller sub-networks, which makes the network easier to organise, safer to run, and less wasteful with addresses. It works by moving the boundary between the network part and the host part of an address.

A subnet mask is a 32-bit number that marks how much of an address is the network and how much is the host. In classful addressing each class has one fixed mask, and subnetting borrows bits from the host side to make more, smaller networks.
| Class | Default subnet mask | Network bits | Host bits |
|---|---|---|---|
| A | 255.0.0.0 | 8 | 24 |
| B | 255.255.0.0 | 16 | 16 |
| C | 255.255.255.0 | 24 | 8 |
Borrowing bits is the practical part. In a Class C network, changing the mask to 255.255.255.192 makes four subnets with fewer hosts each, which suits four small departments better than one big shared network. I borrow bits on paper first before I touch a live mask.
Special IP addresses and their functions
A few addresses do special jobs, and I meet them on every network I check. They are worth knowing by sight, because they explain traces and errors at once.
- Loopback (127.0.0.1): a device talking to itself, used to test the network stack without any cable.
- Broadcast (for example 192.168.1.255): the last address in a network, used to send a message to every host at once.
- Multicast (224.0.0.0 – 239.255.255.255): a message to one group of devices, used for streaming and conferencing.
- APIPA (169.254.x.x): an address a device gives itself when no DHCP server answers, good for local talk only.
The loopback address is the one I test first when a machine cannot reach the network. If 127.0.0.1 answers, the machine’s own stack works, and the fault sits further out.
Challenges of classful addressing
The class system was simple, but its fixed sizes caused five problems that grew with the internet. I list them here because they explain exactly why CIDR replaced it.
- Inefficient use of addresses: a small network got a whole Class C block and a large one a whole Class A block, so space sat unused.
- Limited flexibility: a network could not grow or shrink its block, so it ran out of addresses or sat on too many.
- Address exhaustion: with only about 4.3 billion IPv4 addresses, the waste brought the shortage forward by years.
- Scalability issues: large networks could not fine-tune their size, which made growth painful.
- Inefficient routing: fixed blocks made routing tables larger and harder to manage than they needed to be.
The waste was the real killer. Giving a 50-person office a Class B block with 65,534 addresses is the kind of thing that empties a scarce resource fast.
Address allocation techniques
Addresses reach devices in three ways, and most networks use all three at once. I choose the method by how steady the device needs to be.
- Static allocation: a fixed address typed in by hand, best for servers, printers and routers that must never move.
- Dynamic allocation (DHCP): the router hands out addresses as devices join, which suits laptops and phones.
- Automatic Private IP Addressing (APIPA): a fallback address in the 169.254 range when no DHCP server answers.
Static addresses take more work to manage, and a typo can cause a clash. DHCP is the default at home, and I reserve static addresses only for the devices that truly need them.
Best practices for IP allocation
These habits keep a network tidy as it grows, and they cost nothing to follow.
- Use DHCP for flexibility, and let the router hand out addresses to everyday devices.
- Reserve static addresses for routers, servers and printers, so they never move.
- Plan for growth and leave spare address space for the next batch of devices.
- Subnet for efficiency, so each part of the network is small enough to manage and isolate.
I write the plan down before I change anything. A network with a written address plan is a network that stays fixable.
Security considerations in IPv4 addressing
Private addresses hide the inside of a network from the public internet, and that is a genuine layer of defence. Firewalls, NAT and proper segmentation add more, and they matter most on the border where private meets public.
Public addresses are visible to everyone, so keep the devices behind them patched and locked down. Attacks like DDoS and IP spoofing target exactly those exposed addresses, and a tidy address plan makes them easier to defend. I check for open ports first whenever I look at a public-facing device.
The role of IPv4 classes in network design
Think of the classes as different-sized rooms in a building. Class A is the hall, Class B the office floor, and Class C the single room — each one fits a different size of tenant.
In the early days that was enough, because the internet was small and the rooms were cheap. As networks grew, the fixed rooms wasted space, like giving a whole hall to a family of four, and the shift to classless addressing began. I still use the room analogy when I explain address sizes to new network owners.
Transitioning to classless addressing (CIDR)
CIDR drops the fixed classes and lets a network take exactly the block it needs, no more and no less. It cuts waste, makes growth easy, and shrinks routing tables, which is why it runs the modern internet.
Our CIDR guide works through the notation and the maths, and a CIDR calculator can work out the ranges for you. If classful addressing is the history, CIDR is the tool you actually use today. I reach for the CIDR guide first whenever a network outgrows its class block.
How to tell the class of an IP address
Look at the first number of the address. From 1 to 126 is Class A, 128 to 191 is Class B, 192 to 223 is Class C, 224 to 239 is Class D, and 240 to 255 is Class E.
The number 127 belongs to loopback, so it never counts as a Class A network. I check that first, because it is the one exception people forget.
Classful addressing in practice: what users ran into
192.168.0.x and 192.168.1.x treated as one network
A Cisco community member explained that 192.168.1.0/24 sits inside 192.168.0.0/23. The real mask on your devices decides the network, not the old class letter.
A 10.x address on a home router
Home routers often hand out 10.x addresses with a 255.255.255.0 mask. Class A rules would say /8, but routers use whatever mask is set in DHCP.
Half the range will not answer ping
A MikroTik user could ping 192.168.11.x but not 192.168.10.x on a /23 network. Forum replies said a host with a /24 mask sends the other half to the gateway instead of reaching it directly.










