The network address is the IP address with every host bit forced to zero
The fastest definition is exact: IP address AND subnet mask = network address. The mask preserves the network portion and clears the host portion. You can arrive at the same result with binary AND, CIDR block-size math, or subnet-boundary reasoning. A strong answer should be able to explain why all three methods agree.
Bitwise AND is the mathematical definition
Each IPv4 address is 32 bits. The subnet mask is also 32 bits. In the mask, 1s identify the network portion and 0s identify the host portion. AND keeps a bit only when both inputs contain 1.
Because host-bit positions in the mask are 0, those positions become 0 in the network address.
Example: 192.168.14.88 with 255.255.255.0
A /24 mask keeps the first three octets and clears the fourth. Therefore the network is 192.168.14.0/24.
This is easy because the prefix ends exactly at an octet boundary. The harder cases are prefixes such as /26, /27, /20, and /22 where the boundary lands inside an octet.
Use block size when the partial mask octet is easy to recognize
Find the mask octet that is neither 255 nor 0. Subtract that value from 256. The result is the subnet block size for that octet.
Example: 192.168.14.150/26
A /26 mask is 255.255.255.192. The final-octet block size is 64, so the legal subnet starts are 0, 64, 128, and 192. Address 150 lies in the 128–191 block.
Why 192.168.14.150/26 becomes 192.168.14.128
The fourth octet of 150 is 10010110. The /26 mask's fourth octet is 11000000. AND keeps only the first two network bits in that octet.
The decimal block-size method and binary AND therefore agree.
Example: 10.0.4.77/27
A /27 uses a block size of 32. Boundaries are 0, 32, 64, 96, 128, and so on. Since 77 lies between 64 and 95, the network is 10.0.4.64/27.
Example: 192.168.5.201/28
A /28 uses a 16-address block. The fourth-octet boundaries near 201 are 192, 208, and 224. Because 201 lies between 192 and 207, the network is 192.168.5.192/28.
Example: 172.16.45.200/20
A /20 mask is 255.255.240.0. The interesting octet is the third, not the fourth. Its block size is 16. Third-octet boundaries are 0, 16, 32, 48, 64, and so on. The value 45 belongs to the 32–47 block.
This is one of the most important subnetting habits: do not assume the calculation always happens in the fourth octet.
Example: 172.20.115.8/22
A /22 mask is 255.255.252.0. The third-octet block size is 4. Boundaries include 108, 112, 116, and 120. The value 115 belongs to the 112–115 block.
Use integer block math when you want a repeatable shortcut
For a partial octet, divide the target octet by the block size, discard the remainder, and multiply back by the block size. That gives the beginning of the block.
This is just another way to find the largest legal boundary that does not exceed the host value.
Common prefixes and their block sizes
| Prefix | Mask | Interesting-octet block size | Total addresses |
|---|---|---|---|
| /25 | 255.255.255.128 | 128 | 128 |
| /26 | 255.255.255.192 | 64 | 64 |
| /27 | 255.255.255.224 | 32 | 32 |
| /28 | 255.255.255.240 | 16 | 16 |
| /29 | 255.255.255.248 | 8 | 8 |
| /30 | 255.255.255.252 | 4 | 4 |
Swipe horizontally to see the prefix table.
The network address tells you whether a destination is local
A host compares its own network prefix with the destination's network prefix. If the results match under the same prefix length, the destination is local to the same IP subnet. If the results differ, normal communication requires routing.
Network-address math predicts the host's next Layer 2 target
If the destination is local, the host normally ARPs for the destination's MAC address. If the destination is remote, the host ARPs for the default gateway's MAC address and sends the frame there.
A wrong subnet mask changes the network address—and therefore the forwarding decision
Suppose the intended configuration is 10.10.10.50/24 but the host is mistakenly configured as /16. For destination 10.10.20.8, /24 produces network 10.10.10.0 for the host and 10.10.20.0 for the destination, so the destination is remote. Under /16, both addresses fall under 10.10.0.0/16, so the host incorrectly treats the destination as local.
The interface prefix creates a connected route for the calculated network
When an interface is configured as 192.168.14.150/26, the host or router normally installs an on-link/connected route for 192.168.14.128/26. This route can outrank a default route for matching destinations.
The network address must be the first address in the block
For /27, every network address in the interesting octet must be a multiple of 32: 0, 32, 64, 96, 128, 160, 192, or 224. An answer such as .70 cannot be the network address of a /27 because it is not a legal block boundary.
The original IP must fall inside the network's address block
If you calculate 192.168.5.192/28, that block runs from .192 through .207. Host .201 is inside it, so the answer is plausible. If your calculated network produced a block that did not contain the original host, the math is wrong.
All host bits must be zero in the network address
This is the binary check that never changes. If any host-bit position remains 1, the result is not the network address. This rule is especially useful when working with prefixes outside the memorized /24–/30 range.
Do not round the host up to the next subnet boundary
The network is the beginning of the block containing the host, so you always choose the largest legal subnet boundary that is less than or equal to the host value.
Do not assume every network address ends in .0
That is true for many /24 networks, but not for /25 through /30. A /26 network can start at .0, .64, .128, or .192. A /28 can start every 16 addresses.
Do not assume subnetting always happens in the fourth octet
Prefixes such as /20 and /22 move the partial mask into the third octet. The final octet then becomes entirely host space and becomes 0 in the network address.
Find the network address before opening the answer
10.0.4.77/27
Block size 32. 77 lies in 64–95. Network: 10.0.4.64/27.
192.168.5.201/28
Block size 16. 201 lies in 192–207. Network: 192.168.5.192/28.
172.20.115.8/22
Third-octet block size 4. 115 lies in 112–115. Network: 172.20.112.0/22.
Mixed-prefix challenge
192.168.50.190/26
Block size 64. 190 lies in 128–191. Network: 192.168.50.128/26.
10.50.73.9/21
/21 mask = 255.255.248.0. Third-octet block size 8. 73 lies in 72–79. Network: 10.50.72.0/21.
172.31.199.254/23
/23 mask = 255.255.254.0. Third-octet block size 2. 199 lies in 198–199. Network: 172.31.198.0/23.
Use a subnet calculator to verify reasoning, not replace it
A calculator is valuable after you have predicted the network manually. Compare the calculator's network, broadcast, mask, and host range with your work. If they differ, diagnose the step that failed: prefix conversion, block size, boundary selection, or earlier-octet handling.
Find the network, then predict the forwarding behavior
This is the complete skill: calculate the network address, validate it, then use it to explain what the host should do with a destination.
Example: 192.168.77.190/25
A /25 mask is 255.255.255.128, which divides the final octet into two 128-address blocks: 0–127 and 128–255. Since 190 lies in the second block, the network address is 192.168.77.128/25.
This example is useful because /25 is the first prefix after /24 where the network address may no longer end in .0.
Example: 10.44.8.246/29
A /29 uses mask 255.255.255.248 and an 8-address block size. Nearby boundaries are 232, 240, 248, and 256. Address 246 belongs to 240–247, so the network is 10.44.8.240/29.
Small prefixes are especially good for practicing boundary accuracy because one mistaken block changes the answer quickly.
When the dotted-decimal mask is given, convert the partial octet into a block before doing anything else
Suppose the prompt gives 255.255.248.0 instead of /21. The third octet is the interesting octet. Subtract 248 from 256 to get an 8-value block size. Then use third-octet boundaries 0, 8, 16, 24, and so on. This keeps dotted-decimal questions from feeling like a different skill than CIDR questions—they are the same calculation expressed in another notation.
Do not confuse the first address of the block with the last address of the block
The network address is the beginning of the subnet. The broadcast address is the final address in a conventional IPv4 subnet. Once you know one boundary and the block size, you can derive the other—but they serve different purposes.
If a question asks only for the network address, stop at the first boundary. The broadcast calculation is useful as a validation step, not a replacement answer.
Use a four-line scratch process under time pressure
For 10.50.73.9/21, write: /21 → mask 255.255.248.0 → block 8 → boundaries 64, 72, 80 → network 10.50.72.0. This routine is short enough to use quickly but still leaves an auditable trail if you need to check your work.
24 Network Address digital flashcards
Practice bitwise AND, CIDR block sizes, /24–/30 boundaries, earlier-octet subnetting, validation, local-vs.-remote behavior, wrong-mask effects, and route implications.