CIDR and IPv4 Subnetting: Worked Examples for Developers
Reviewed September 3, 2026 · Maintained by William
CIDR notation combines an IPv4 address with a prefix length. The prefix says how many leading bits identify the network.
/24 example
192.168.10.25/24 uses mask 255.255.255.0. The network is 192.168.10.0 and the traditional broadcast address is 192.168.10.255.
/26 example
A /26 leaves 6 host bits, so each block contains 64 addresses. The last octet boundaries are 0, 64, 128 and 192.
192.168.10.70/26
network: 192.168.10.64
broadcast: 192.168.10.127
range: 192.168.10.64 - 192.168.10.127In a traditional subnet, addresses between network and broadcast are host addresses, but provider-specific reservations can reduce assignable counts.
/30 example
A /30 has four addresses per block. For 10.0.0.6/30, the block is 10.0.0.4–10.0.0.7, with the traditional usable pair .5 and .6.
/31 and /32 are edge cases
A simplistic “subtract network and broadcast” rule breaks down for modern point-to-point /31 use described by RFC 3021. A /32 identifies one address. Calculators should show the arithmetic and avoid overgeneralizing “usable hosts.”
What subnet math cannot tell you
- Which routes your router will announce.
- Which addresses a cloud provider reserves.
- Whether NAT, firewall rules or VLANs make the address reachable.
- Whether an address is appropriate for public use.
The CIDR Range Calculator and IP Subnet Calculator focus on standard IPv4 arithmetic and explicitly separate that from routing policy.
References
Worked /24 example
For 192.168.20.42/24, the first 24 bits identify the network. The mask is 255.255.255.0, so the network is 192.168.20.0 and the broadcast address is 192.168.20.255. In traditional host-addressing terms, the range between them is 192.168.20.1 through 192.168.20.254.
Worked /26 example
A /26 leaves 6 host bits, giving blocks of 64 addresses. In the last octet the block boundaries are 0, 64, 128 and 192.
Input: 192.168.10.77/26
Block: 192.168.10.64 - 192.168.10.127
Network: 192.168.10.64
Broadcast: 192.168.10.127
Traditional host range: 192.168.10.65 - 192.168.10.126The useful mental shortcut is block size, but binary reasoning is more general: mask off the host bits for the network and set host bits to 1 for the broadcast address.
Worked /30 example
A /30 has four addresses per block. 10.0.0.9/30 falls in the block 10.0.0.8–10.0.0.11. Traditionally, .8 is network, .11 broadcast and .9–.10 are host addresses. /30 is therefore common in older point-to-point subnetting examples.
/31 and /32 are special cases
Do not mechanically subtract two usable addresses for every prefix. RFC 3021 defines /31 use for point-to-point links, where both addresses can be endpoints and the traditional network/broadcast distinction is not used the same way. A /32 identifies one IPv4 address. Calculators should surface these caveats rather than applying a classroom formula blindly.
Cloud “usable address” counts can differ
Cloud providers may reserve addresses inside a subnet for platform functions. CIDR math describes the address block; provider policy describes what you may assign. When planning capacity, use the provider’s current subnet-reservation documentation in addition to the calculated range.
Subnet debugging checklist
- Write the prefix and dotted mask explicitly.
- Calculate the containing block boundary, not just the host count.
- Check whether the target IP is actually inside the expected range.
- Separate subnet arithmetic from routing, firewall, VLAN and NAT policy.
- For /31-/32, apply the intended operational use rather than generic “network/broadcast” assumptions.
Use the CIDR Range Calculator or IP Subnet Calculator to check arithmetic, then validate routing and platform reservations in the environment where the network will exist.