Subnet Masks Explained: Dotted-Decimal and CIDR Notation

Networking · Last updated · About 5 min read

What a subnet mask does and the two ways to write it, dotted-decimal (255.255.255.0) and CIDR slash notation (/24), with a conversion table, worked examples and common mask mistakes.

A subnet mask tells a device which part of an IP address identifies the network and which part identifies the device on that network. It is how a computer decides whether another address is "local" (reach it directly) or "somewhere else" (send it to the router). Every device with an IP address also has a subnet mask. There are two common ways to write the same mask, and they mean exactly the same thing.

The two ways to write a subnet mask

Notation Example Used where
Dotted-decimal (also "netmask") 255.255.255.0 Windows network settings, DHCP server scopes, and the network screens of most routers, gateways and radio equipment
CIDR / prefix length (slash notation) /24 Router configurations, firewall rules, documentation, cloud networking, addresses written as 192.168.1.0/24

They are two spellings of one fact: the number after the slash is the count of leading 1 bits in the mask. The mask 255.255.255.0 is 24 ones followed by 8 zeros, so it is /24.

/24  =  11111111.11111111.11111111.00000000  =  255.255.255.0
/26  =  11111111.11111111.11111111.11000000  =  255.255.255.192

CIDR (Classless Inter-Domain Routing, 1993) replaced the old fixed "class A, B, C" network sizes with prefixes of any length (networking-ip-address-classes). A third notation, the wildcard mask used in some router access lists, is simply the mask inverted (0.0.0.255 for a /24); see acl-access-control-list.

How the mask is used

The device combines the address with the mask bit by bit (a logical AND). Wherever the mask has a 1, the address bit belongs to the network; wherever it has a 0, the bit identifies the host.

A 32-bit address split by a /26 subnet mask The address 192.168.1.130 is shown as 32 binary bits in four octets of eight. The first 26 bits are marked as network bits and the last 6 as host bits, matching the mask 255.255.255.192, whose bits are 26 ones followed by 6 zeros. Six host bits give 64 addresses: network address 192.168.1.128, usable hosts 192.168.1.129 through 192.168.1.190, and broadcast address 192.168.1.191. What the mask actually marks192.168.1.130 with mask 255.255.255.192, written /26 because 26 bits are networkAddressMask11110101010101011101110111010101010101010101011111010000000010001922551682551255130192the /26 boundary26 network bits — identical on every device in this subnet6 host bits — 2⁶ = 64 addresses, of which 62 are usableWhat that gives youNetwork address192.168.1.128all host bits 0 — names the subnetFirst usable host192.168.1.129Last usable host192.168.1.19062 addresses in betweenBroadcast address192.168.1.191all host bits 1 — reaches everyoneChange the mask and only the boundary moves. A /24 would put the whole last octet in orange: 254 usableaddresses instead of 62.
The mask is simply a line drawn through the 32 bits. Everything to its left says which network; everything to its right says which device.

Example: a PC is 192.168.1.130 with mask 255.255.255.192 (/26).

Address   192.168.1.130  = 11000000.10101000.00000001.10000010
Mask      255.255.255.192= 11111111.11111111.11111111.11000000
Network   192.168.1.128  = 11000000.10101000.00000001.10000000

The network is 192.168.1.128/26. The host part is the remaining 6 bits, which allow 64 addresses in this subnet. Two of them are reserved: the network address (all host bits 0, here .128) and the broadcast address (all host bits 1, here .191). That leaves 62 usable host addresses, .129 to .190. See networking-unicast-broadcast-multicast.

Common masks

CIDR Dotted-decimal mask Total addresses Usable hosts Typical use
/8 255.0.0.0 16,777,216 16,777,214 Very large private network (10.0.0.0)
/16 255.255.0.0 65,536 65,534 Large site or campus
/24 255.255.255.0 256 254 The common default for a LAN
/25 255.255.255.128 128 126 Half a /24
/26 255.255.255.192 64 62 Quarter of a /24
/27 255.255.255.224 32 30 Small segment
/28 255.255.255.240 16 14 Small equipment group
/29 255.255.255.248 8 6 A few devices
/30 255.255.255.252 4 2 Point-to-point link between two routers

The usable hosts are always the total minus 2. (A /31 with two addresses is allowed on point-to-point links by modern equipment.)

The mask numbers in the last octet of the table (128, 192, 224, 240, 248, 252) are all you need to memorize. The block size of each subnet is 256 minus that number: a mask ending in .192 gives blocks of 64 (.0, .64, .128, .192).

Worked examples

Which subnet is 10.5.20.77/26 in? Block size is 64, so the blocks start at 0, 64, 128 and 192. 77 falls in the 64 block: network 10.5.20.64, broadcast 10.5.20.127, usable .65 to .126.

Why can't these two talk directly? One PC is 192.168.1.10 and another is 192.168.2.10, both with mask 255.255.255.0. The network parts (192.168.1 and 192.168.2) differ, so each sees the other as "somewhere else" and sends the traffic to its default gateway. With no router that connects them, nothing gets through. If both had mask 255.255.0.0 (/16) they would be in the same subnet and talk directly.

How many addresses does a radio site need? Count every device that needs an address (routers, gateways, repeaters, consoles, recorders, printers), add room for growth, and pick the smallest block that fits. A site with 20 devices that may grow to 28 fits a /27 (30 usable); 40 devices needs a /26.

What happens with the wrong mask

  • Too long (too small) a mask on one device. For example /26 on a PC that should be /24. The PC treats some devices on its real network as "somewhere else", so it can reach some neighbors but not others. This is the classic "I can ping one device but not another on the same switch" symptom.
  • Too short (too big) a mask. The device believes distant addresses are local and tries to reach them directly instead of through the router, so they are unreachable.
  • Different masks on devices in the same subnet. They disagree about the subnet's boundaries, with confusing and one-directional failures.
  • The network or broadcast address given to a device. The first and last address of a subnet (for example .0 and .255 in a /24) cannot be used by a host.

When troubleshooting, run ipconfig /all (see dhcp-vs-static-ip-addressing) on both devices, compare the mask, and work out each address's network as above. network-troubleshooting-commands lists the other tools.

Why subnet at all

Dividing a network into smaller subnets keeps broadcast traffic local, makes it practical to separate radio-system infrastructure from general office traffic (ip-networking-in-two-way-radio-systems), and lets routers and firewalls apply different rules to different groups (networking-switches-routers-hubs, networking-vlan). For the addressing scheme itself see networking-ipv4.

This article is part of the CTI Products Knowledge Base. Questions about RadioPro or MCN products? Call +1 (513) 595-5900 or email info@ctiproducts.com.

Informational only. This article is general background and is not engineering, safety, legal or regulatory advice. Verify requirements with the equipment manufacturer and the applicable authority before relying on it. Product names and trademarks belong to their respective owners.