A technician opens a wall outlet to add a drop and finds white/green on pin 1. Every other jack in the building is white/orange. Nothing is broken, nothing is labeled, and the last person who knew which standard the site uses left years ago.
That is the whole problem in one outlet. T568A and T568B are not a technical disagreement — they are a documentation failure that compounds.
The Actual Difference, in One Table
Both standards wire an eight-position modular plug straight through — pin 1 to pin 1, pin 8 to pin 8. The two pinouts differ in exactly one place: the orange and green pairs are interchanged. Pairs 2 and 3 swap. Nothing else moves.
Pin | T568A pair | T568B pair | T568A color | T568B color |
|---|
1 | 3 | 2 | white/green | white/orange |
|---|
2 | 3 | 2 | green | orange |
|---|
3 | 2 | 3 | white/orange | white/green |
|---|
4 | 1 | 1 | blue | blue |
|---|
5 | 1 | 1 | white/blue | white/blue |
|---|
6 | 2 | 3 | orange | green |
|---|
7 | 4 | 4 | white/brown | white/brown |
|---|
8 | 4 | 4 | brown | brown |
|---|
Two things in that table matter later.
First, pins 1 and 2 form a pair, as do 3 and 6 — not 3 and 4. The pairs are deliberately not adjacent on the connector, and the reason is that the original registered-jack layout would have separated the outer pairs too far to meet the electrical requirements of high-speed LAN protocols.
Second, pins 4, 5, 7 and 8 are identical in both standards, and pair 1 sits on the center pins in T568A, T568B and USOC alike.
Why Two Standards Exist at All
This is not a case of one standard correcting the other. Both are allowed under ANSI/TIA-568.2-D, and the split is historical.
T568A descends from USOC, the telephone wiring scheme. Fluke Networks describes it as the recognized preferred wiring pattern of the standard, because it provides backward compatibility to both one-pair and two-pair USOC wiring. The U.S. government requires T568A for wiring done under federal contracts.
T568B matches the older AT&T 258A color code. In the 1990s, when the original TIA/EIA-568 was published, the most widely installed pattern in UTP infrastructure was AT&T 258A — which is why the standard included that same pattern as a secondary option for existing installations. Many organizations still use T568B out of inertia.
Worth knowing before a project meeting: Australia typically uses T568A while US enterprises prefer T568B, even where federal facilities require the opposite.
Electrically Identical, Operationally Not
Fluke’s position is blunt: from a technical standpoint there is no difference between T568A and T568B. Same transmission performance, same protocol support including Gigabit Ethernet. NetAlly says the same from the other direction — the distinction is purely wiring arrangement.
So the choice does not affect speed or reliability. What it affects is whether the next person can work on the building. Fluke’s warning: the two are not interchangeable, and wiring should match color to color and stripe to stripe.
Read that carefully, because it is the crux. Equivalent means either one works. Interchangeable would mean you can use both in the same building — and you cannot, not because performance suffers broadly, but because of one specific case.
The Case Where It Actually Breaks
Here is where the two swapped pairs stop being cosmetic. 10BASE-T and 100BASE-TX use only two pairs — pins 1 and 2 for transmit, pins 3 and 6 for receive. Those are precisely the two pairs on which T568A and T568B differ.
A cable terminated T568A on one end and T568B on the other is therefore a crossover cable: the transmit and receive pairs arrive swapped. That is not an analogy — it is the same wires in the same positions as a deliberately built crossover.
The consequence splits by speed:
- On 10BASE-T and 100BASE-TX, an A-to-B cable crosses transmit to receive. That was occasionally useful historically, and today auto-MDIX on most equipment corrects it — which is why the fault often goes unnoticed until it reaches hardware that does not.
- On 1000BASE-T and above, all four pairs carry data simultaneously, in both directions, as bidirectional signals. There is no dedicated transmit pair to swap, so the same cable that crosses 100BASE-TX does not cross Gigabit in the same way.
A 100 Mbps link and a Gigabit link behaving differently on the *same* bad cable is what makes A/B faults so hard to diagnose. The link works, then it does not, and the only variable that changed was the switch port.
There is a second, quieter cost that shows up in test results rather than in connectivity. Swapping two wires between different pairs causes crosstalk, defeating one of the reasons the wires are twisted in pairs at all. A miswired link can pass continuity and still carry a real performance penalty.
Why PoE Survives a Standard Mismatch
PoE comes in two modes. Mode A delivers power on the data pairs of 100BASE-TX or 10BASE-T. Mode B delivers power on the spare pairs. In Mode A, pins 1 and 2 form one side of the 48 VDC supply and pins 3 and 6 the other. In Mode B, pins 4 and 5 form one side and pins 7 and 8 the return — which is why Mode B requires a four-pair cable.
Now add the pinout. In T568A, pins 1 and 2 are the green pair; in T568B, the orange pair. Under Mode A, the supply polarity therefore lands on a different colored pair depending on which standard the cable was terminated to.
That sounds like it should matter. It does not — by design. PoE polarity is deliberately free, and that is what allows the technology to accommodate crossover cables, patch cables and auto-MDIX. A standard mismatch that breaks a 100BASE-TX data link will generally still pass power.
Two rules follow. The power sourcing equipment, not the powered device, decides whether Mode A or Mode B is used, and devices implementing only one mode are disallowed by the standard. And on 1000BASE-T there are no spare pairs, so power is delivered by the phantom technique rather than a dedicated pair.
A practical warning follows: because PoE keeps working, a miswired link can power an access point for years while carrying degraded data, and nobody investigates.
The Telephone Legacy Nobody Expects
If the building ever carried analog phone lines, A and B stop being equivalent in a way that produces real faults.
Both standards put pair 1 on the center pins, 4 and 5, which is why a single phone line works on either. But line 2 lands on pins 3 and 6 — pair 2 in USOC and T568A jacks, pair 3 in T568B jacks.
A second line through a T568B outlet therefore connects to the wrong pair, and USOC plugs cannot reach pair 3 or 4 from a T568A jack, or pair 2 or 4 from a T568B jack, without splitting pairs. Splitting a pair means the two conductors of a signal are no longer twisted together — the classic source of hum and noise on a line that passed continuity.
Finding Out What You Already Have
You cannot tell A from B from the outside of a jack, and you should not trust the label on an old patch panel.
Look at the conductors at a termination — pin 1 is the giveaway. White/green means A; white/orange means B. Check more than one outlet, because inherited buildings are often a mix, usually where a later contractor added a wing.
One complication before you open a jack: the wiring at the rear varies by manufacturer and may not follow the front sequence, usually because a small circuit board inside routes the connections. Terminate by the color code on the jack’s own label, not the rear terminal positions.
A wire map test is the other half: a proper tester identifies exactly which pins are misconnected, and selecting the right standard is what makes the report meaningful. Run against the wrong standard, it flags every correctly terminated jack in the building.
How to Decide
The rule is easier than the debate suggests.
Match what is already installed. The standard’s allowance for T568B exists for exactly this reason: adding to an installation that used that pattern. A new wing that does not match the rest of the building is a fault generator, not an upgrade.
On new construction with no legacy, it is a policy decision, not a technical one. T568A carries the standard’s recommendation, the federal-contract requirement, and backward compatibility to both one-pair and two-pair USOC. T568B matches the color code most equipment manufacturers use.
Then write it down. Put the standard and the date in the cabling documentation. The cost of this decision is entirely in the second visit — the one where nobody knows. Our patch cord range, and the Cat5e and Cat6 premise cables it terminates into, are built and labeled to the standard the customer specifies. Every assembly goes through test and inspection, with the certifications we hold listed by standard. If inherited wiring is causing test failures and the standard is unclear, send us the wire map report and we will tell you which termination pattern it describes.
Frequently Asked Questions
Does T568A or T568B affect network speed?
No. Both have identical transmission performance and support the same protocols, including Gigabit Ethernet. The difference is which colored pair lands on which pin.
What happens if I terminate one end A and the other end B?
You have built a crossover cable. On 10BASE-T and 100BASE-TX the transmit and receive pairs are swapped, because those are the two pairs the standards differ on. On 1000BASE-T all four pairs carry data bidirectionally, so the failure mode differs.
Will PoE still work if the standard is mismatched?
Usually yes. PoE polarity is deliberately free to accommodate crossover and patch cables. The power sourcing equipment chooses Mode A or Mode B, and a device limited to a single mode is not allowed by the standard.
Can I mix them in the same building?
Do not mix. Both are allowed under ANSI/TIA-568.2-D and either is correct alone. A building that uses both produces faults that look like hardware failures and cost hours to trace.
The Short Version
T568A and T568B are the same wiring with pairs 2 and 3 exchanged — orange and green. Both are permitted by ANSI/TIA-568.2-D and both perform identically. T568A is the standard’s preferred pattern, descends from USOC telephone wiring, and is required on U.S. federal contracts. T568B matches the older AT&T 258A code and has been the more widely installed scheme for decades.
They are equivalent but not interchangeable. The difference lands on exactly the two pairs used by 10BASE-T and 100BASE-TX, which is why an A-to-B cable is a crossover cable and why a 100 Mbps link breaks while a Gigabit link on the same cable does not.
PoE survives a mismatch because its polarity is deliberately free — which is why a miswired drop can power a device for years while carrying degraded data. And where analog phone lines share the cable, line 2 lands on a different pair depending on the standard.
Pick the standard, match the building, and write it into the documentation. The wiring will outlast the memory of whoever chose it.
Sources
- Fluke Networks — Differences Between Wiring Codes T568A vs T568B, covering the interchange of the orange and green pairs, both standards being allowed under ANSI/TIA-568.2-D, T568A as the recognized preferred pattern with backward compatibility to one-pair and two-pair USOC and the U.S. federal contract requirement, T568B matching the AT&T 258A color code with only single-pair USOC compatibility, the pin assignment detail for both standards, the statement that there is no technical difference in transmission performance or protocol support, the statement that the two are not interchangeable and that wiring should match color to color and stripe to stripe, T568B being more common in home networks for matching equipment manufacturer color codes, and the note that rear jack wiring varies by manufacturer and is usually routed by a small printed circuit board — https://www.flukenetworks.com/knowledge-base/application-or-standards-articles-copper/differences-between-wiring-codes-t568a-vs
- Wikipedia — ANSI/TIA-568, covering the revision lineage from the 1991 initial issue through revisions A, B, C, D and the current E, the pin-to-pair tables for T568A and T568B, the statement that pairs 2 and 3 are exchanged, the pair ordering of blue, orange, green and brown, the straight-through nature of both pinouts, the pin pairing of 1-2, 3-6, 4-5 and 7-8, the interchangeability of either configuration within one installation when both ends match, the definition of an A-to-B cable as a crossover cable because the pairs used by older two-pair Ethernet standards are exactly the pairs on which the standards differ, the crosstalk consequence of swapping wires between pairs, the 10BASE-T and 100BASE-TX transmit and receive pin usage against 1000BASE-T signal identifiers on all four pairs, the USOC telephone line 1 and line 2 behavior on pins 4-5 and 3-6 and the confusion this creates with T568B, the split-pair consequence for USOC plugs, the use of pairs 1 and 3 in T568A for T1 service, the 90 meter permanent link and 100 meter channel with a maximum 5 meter patch cord, the category bandwidths including 5e, 6, 6A and 8 at 2,000 MHz, that Categories 7 and 7A were not officially recognized by TIA, and the note that 8P8C connectors are often incorrectly referred to as RJ45 — https://en.wikipedia.org/wiki/ANSI/TIA-568
- EtherWAN — What is the PoE power pins assignment mode A & B?, covering Mode A delivering power on the data pairs of 100BASE-TX or 10BASE-T and Mode B on the spare pairs, the phantom technique on 1000BASE-T where there are no spare pairs, pins 1 and 2 forming one side of the 48 VDC and pins 3 and 6 the other in Mode A, pins 4-5 and 7-8 in Mode B and the four-pair requirement, the free polarity allowing PoE to accommodate crossover cables, patch cables and auto-MDIX, and the PSE rather than the PD deciding the mode with single-mode-only devices disallowed by the standard — https://www.etherwan.com/us/support/faq/what-poe-power-pins-assignment-mode-b
- NetAlly — T568A vs T568B: Breaking Down Ethernet Wiring Standards, covering the green and orange pair swap, the full pin-by-pin wire order for both standards, the statement that there is no performance difference and that the distinction is purely wiring arrangement, the historical derivation of T568A from USOC and T568B from AT&T 258A, Australia typically using T568A against US enterprise preference for T568B and US federal facilities requiring T568A, the guidance to check both country and industry standards, and the tester workflow of selecting the wiring standard so that miswiring is reported against the correct pin reference — https://www.netally.com/tech-tips/t568a-vs-t568b/
