Quick answer: Cat5 and Cat5e are both rated to 100 MHz, so the headline number is identical. What changed is the limit on crosstalk: Cat5e specifies far-end crosstalk and return loss tightly enough for 1000BASE-T over the full 100 m, while original Cat5 was designed for 100BASE-TX and often works at gigabit only by luck. Cat5e also supports 2.5GBASE-T at 100 m. If the cable is already in the wall and passes a test, keep it; if you are buying, there is no reason to buy Cat5.
The difference is not visible: same four twisted pairs, same 24 or 23 AWG conductors, same RJ45 plugs. It lives in the test report, in how much signal leaks from one pair into another at 100 MHz, and in whether the link can run all four pairs simultaneously in both directions without the receiver’s echo canceller giving up.
What actually changed
Same bandwidth, stricter limits
Both categories specify 100 MHz. Cat5e added limits for power sum near-end crosstalk, far-end crosstalk and return loss, which are the parameters that matter when gigabit uses all four pairs at once. A cable that meets Cat5 limits can pass 100 Mbit/s comfortably and may fail gigabit on a long run or in a tight bundle.
Why gigabit needs the tighter limits
1000BASE-T transmits on all four pairs in both directions simultaneously. Each receiver must cancel the echo of its own transmission and the crosstalk from the other three pairs. Cat5e guarantees those impairments stay inside the canceller’s range at 100 m; Cat5 does not, which is why the same installation can link at gigabit on one port and fall back to 100 Mbit/s on another.
Speed and distance by category
| Category | Bandwidth | Rated speed at 100 m | Realistic in practice |
|---|
| Cat5 | 100 MHz | 100BASE-TX | Gigabit often works, not guaranteed |
| Cat5e | 100 MHz | 1000BASE-T, 2.5GBASE-T | Gigabit reliable, 2.5G on clean runs |
| Cat6 | 250 MHz | 5GBASE-T | 10G to 55 m |
| Cat6a | 500 MHz | 10GBASE-T | 10G to 100 m, shielded |
Installation decides more than the label
Solid versus stranded
Permanent cable uses solid conductors, which have lower loss and are not meant to flex. Patch cords use stranded conductors for flexibility and lose noticeably more per metre. A 90 m permanent link made of stranded patch cable will fail where the same length of solid cable passes, and the label on the box will not warn you.
Termination and untwisting
Keep the untwisted length under 13 mm at each jack. Crosstalk is generated where the pairs run parallel, so a sloppy termination on Cat5e cable can produce a link worse than a careful one on Cat5. Use the same wiring pattern at both ends and test the permanent link rather than only checking that it links.
Reuse or replace
If the existing cable is Cat5e and the requirement is gigabit or 2.5G, reuse it after testing. If it is unmarked Cat5 and the run is over 60 m or bundled with power, expect problems and plan a replacement. For new work, Cat6 costs little more and leaves headroom for 5G; the Cat6 LAN cable overview covers the construction differences, and the Cat5 vs Cat6 vs Cat6a comparison maps the categories to speeds.
For patch cords, choose Cat5e only if the run is short and the switch is gigabit. A Cat5e cable in good condition is electrically fine for that job; the failure mode is mechanical, in the plug latch and the strain relief, not in the copper.
Testing what is already installed
A cheap continuity tester confirms that eight conductors are connected, which is not the same as a link that will carry gigabit. A certifier measures insertion loss, near-end and far-end crosstalk and return loss against the category limits, and reports a pass or fail per parameter. For an existing installation, test a sample of the longest and most bundled runs first: those are the ones most likely to fail, and their result tells you whether the rest of the cabling is worth keeping or should be replaced during the same visit.
Record the results. A certification report per outlet, with the length and the margin to the limit, is the document that settles an argument about whether a fault is the cabling or the switch. It also shows which links are closest to failing, so a future upgrade to 2.5G or 5G can begin with the ones that have the most margin instead of testing everything again.
Practical scenarios
- Office with existing Cat5 runs: test each link at 100 MHz; those that pass can carry gigabit, and the ones that fail usually need re-termination rather than new cable.
- Home with 2.5G switches: Cat5e is enough for 100 m, but keep the run away from mains cable and avoid flat cable, which is not rated to the category.
- New build: install Cat6 or Cat6a once; the labour dominates the cost and re-pulling later costs more than the cable.
- PoE camera on an old Cat5 run: check conductor gauge before assuming it is fine, because thin 26 AWG patch cable in the middle of the run causes voltage drop.
FAQ
Is Cat5 the same as Cat5e?
No. Both are 100 MHz cables, but Cat5e has stricter crosstalk and return loss limits, which is what makes gigabit dependable over 100 m. Cat5 was specified for 100BASE-TX and often runs gigabit, but that is outside its rating.
Can Cat5e run 10 gigabit?
Not to 100 m. 10GBASE-T needs 500 MHz, so Cat6a is the minimum for full distance and Cat6 manages about 55 m. Cat5e will negotiate 2.5GBASE-T at 100 m on a clean run, which is the realistic upgrade path.
Does Cat5e need shielding?
No, for normal office and home installations. Shielded cable only helps when the run is in a high-interference environment and the shield is properly bonded at both ends. A floating shield on an unshielded system makes crosstalk and pickup worse, not better.
