Network Cable Color Code: T568A, T568B, Crossover, and Rollover Pinouts Explained

When a customer says an Ethernet cable is wired correctly because the lights came on, I still want the pair map and pin order checked against T568A or T568B. At 1000BASE-T, all 4 pairs are active, so a cable that looks acceptable on a basic continuity test can still fail certification if pair integrity or split-pair control is wrong.

Many engineers search for network cable color code because they need a fast answer at the bench: which wire goes to pin 1, what changes between T568A and T568B, and when do you actually need a crossover cable instead of a straight-through patch lead? In manufacturing, though, the question is broader. Color code affects harness documentation, operator training, outgoing inspection, and field-service recovery. If the assembly drawing, work instruction, and test fixture do not all reference the same pinout standard, you end up with avoidable rework and intermittent network faults.

On a site like YourPCB, this matters because Ethernet is rarely a standalone commodity cable. It is often part of a larger box-build or interconnect program that includes custom length control, shield termination, overmolding, panel I/O, and PCB-connected subsystems. A correct pinout is the starting point, not the whole job.

This guide explains the standard network cable color code sequences, where T568A and T568B differ, when crossover and rollover cables still matter, and how to prevent the manufacturing mistakes that create failed link tests. If you need adjacent background, review our wire connector types guide, how to crimp wires guide, types of power connectors guide, and wire harness contract manufacturing service before releasing a custom Ethernet assembly.

What Is a Network Cable Color Code?

A network cable color code is the prescribed conductor order used when terminating twisted-pair Ethernet cable into an 8-position modular connector, commonly called 8P8C and often informally called RJ45. The code is not arbitrary decoration. It preserves the intended twisted-pair grouping so the cable maintains impedance stability, crosstalk control, and predictable transmission behavior.

For background and standards context, TIA/EIA-568, Ethernet over twisted pair, 8P8C modular connector, and Power over Ethernet are useful references because the color sequence is directly tied to pair assignment and application class.

In practice, the right network cable termination has to satisfy 5 conditions at once:

1. Match the intended pinout standard exactly, usually T568A or T568B.
2. Keep each twisted pair on the correct pin pair assignment rather than creating split pairs.
3. Maintain untwist length short enough to preserve return loss and NEXT margin.
4. Fit the cable category, shield design, and strain-relief hardware being used.
5. Pass the correct test method, from wiremap to certification, before shipment.

Network Cable Color Code Chart

The fastest way to avoid mistakes is to compare the common Ethernet termination patterns side by side.

Cable or standard	Pin 1	Pin 2	Pin 3	Pin 4	Pin 5	Pin 6	Pin 7	Pin 8
T568A straight-through	White/Green	Green	White/Orange	Blue	White/Blue	Orange	White/Brown	Brown
T568B straight-through	White/Orange	Orange	White/Green	Blue	White/Blue	Green	White/Brown	Brown
Crossover end A	White/Orange	Orange	White/Green	Blue	White/Blue	Green	White/Brown	Brown
Crossover end B	White/Green	Green	White/Orange	Blue	White/Blue	Orange	White/Brown	Brown
Rollover end A	White/Orange	Orange	White/Green	Blue	White/Blue	Green	White/Brown	Brown
Rollover end B	Brown	White/Brown	Green	White/Blue	Blue	White/Green	Orange	White/Orange

That table shows the key fact: T568A and T568B differ only in the placement of the green and orange pairs. The blue pair stays on pins 4 and 5, and the brown pair stays on pins 7 and 8.

1. T568A Color Code

T568A assigns the green pair to pins 1 and 2 and the orange pair to pins 3 and 6. It remains a recognized standard and is still used in some structured cabling environments, government work, and legacy installations where T568A was specified across the building.

The practical rule is simple: T568A is fully valid if both ends and the installation documentation are aligned. The problem is not that T568A is wrong. The problem is mixing T568A and T568B unintentionally inside the same production lot.

If your customer drawing, patch panel labeling, or site standard calls for T568A, follow it exactly and document it on the traveler, test report, and package label. Consistency matters more than personal preference.

2. T568B Color Code

T568B assigns the orange pair to pins 1 and 2 and the green pair to pins 3 and 6. In many commercial environments it became the more familiar field convention, which is why many bench diagrams default to T568B.

Electrically, T568B is not faster than T568A. A compliant Category 5e, 6, or 6A link does not become better just because one scheme is more common in your region. The reason assemblers often choose T568B is simply to match installed plant expectations and reduce field confusion during moves, adds, and changes.

For custom Ethernet harnesses we usually lock the pinout in the work instruction header and on the first-article label. That sounds basic, but it prevents one of the most common low-volume failures: 10 pieces built as T568B, 2 emergency replacements built as T568A, and hours wasted proving the switch is not the problem.

3. Straight-Through vs Crossover: What Actually Changes?

A straight-through Ethernet cable uses the same standard at both ends, either T568A-to-T568A or T568B-to-T568B. That is the normal choice for patch cords, switch-to-device links, and most modern industrial Ethernet harnesses.

A crossover cable deliberately swaps the transmit and receive pairs by terminating one end as T568A and the other as T568B. Historically this was needed when connecting similar devices directly, such as switch-to-switch or PC-to-PC without auto-MDI/MDIX support.

Today crossover cables are less common because many interfaces can automatically detect and correct pair orientation. But they still appear in legacy support kits, maintenance environments, and special embedded systems. If you build one, label it clearly. An unlabeled crossover cable can waste 30 minutes of troubleshooting faster than almost any other low-cost interconnect error.

4. Rollover Cable Color Code

A rollover cable is different from a crossover cable. It reverses the pin order end to end so pin 1 goes to pin 8, pin 2 to pin 7, and so on. It is commonly associated with console-port access on network equipment rather than standard Ethernet data links.

This distinction matters because field techs often describe any "non-standard Ethernet cable" as a crossover cable, which is wrong. If the cable is intended for console access, write rollover explicitly into the drawing and mark the label accordingly.

In production, rollover cables should never be mixed into ordinary patch-cord stock. They belong in separate inventory with distinct labeling, because the connector looks familiar even though the function is completely different.

5. The Real Failure Mode: Split Pairs

The most dangerous mistake is not choosing T568A instead of T568B. It is creating a split pair. A split pair happens when continuity looks correct pin to pin, but the conductors that form a twisted pair are separated and re-matched incorrectly across the connector.

For example, a cable may still show 1-to-1 continuity on all 8 positions, yet fail because pins 1 and 2 are no longer a real pair from the same twist. That destroys the cable's differential performance. At 10BASE-T or 100BASE-TX you may see intermittent operation. At 1000BASE-T, where all 4 pairs are active, the margin can collapse completely.

That is why a simple continuity buzzer is not enough for production release. At minimum, use a tester with wiremap and split-pair detection. For higher-category assemblies or customer-specified links, use a certification-grade tester matched to the target category.

6. How To Terminate a Network Cable Correctly

A repeatable Ethernet termination process usually follows 6 steps.

1. Confirm the required standard before stripping

Do not let operators decide from memory. The traveler should say T568A, T568B, crossover, or rollover explicitly. If shielded cable is required, specify the connector body and shield-bond method too.

2. Keep pair twist as close as possible to the contacts

Untwisting too much conductor length hurts electrical performance. As a practical shop-floor rule, keep the untwisted section to about 13 mm or less unless the connector system specifically requires otherwise.

3. Preserve jacket capture and strain relief

The cable jacket should extend into the plug enough for the crimp tab or strain feature to hold the jacket, not only the insulated conductors. This matters for pull strength and long-term reliability.

4. Use the right plug for conductor style

Solid conductors and stranded conductors often need different plug designs. Mixing them may pass initial continuity and then fail after a few insertions or minor flexing.

5. Test more than continuity

For prototype runs, continuity and wiremap may be acceptable if the customer agrees. For production or category-rated assemblies, add split-pair detection and, where required, certification for insertion loss, return loss, and NEXT.

6. Label special cables clearly

Crossover, rollover, PoE-only, shielded drain-bonded, or custom-pinned cables should never leave the line with generic patch-cord labeling.

Common Manufacturing Mistakes With Ethernet Cable Assemblies

The first mistake is treating all modular plugs as interchangeable. Cat5e, Cat6, Cat6A, solid, stranded, shielded, and pass-through connector designs do not all build the same way.

The second mistake is losing pair discipline during hand dressing. Operators flatten the conductors for convenience, but once pair identity is lost, split-pair defects become much more likely.

The third mistake is failing to control shield termination. On shielded cable, 360-degree shield continuity and drain management affect EMC performance just as much as the pinout affects link integrity. If the application includes noisy drives, servo systems, or industrial enclosures, shield termination deserves the same process control as conductor order.

The fourth mistake is forgetting the system context. A custom Ethernet lead may need panel feedthroughs, PCB-connected magnetics, strain-relieved I/O, or combined harness routing inside a larger build. That is why our low volume wire harness assembly service, bespoke cable manufacturers service, and obsolete connector replacement service often get involved in Ethernet-adjacent programs instead of treating them as off-the-shelf patch cords.

If the cable will carry PoE above 30 W, I want conductor gauge, bundle temperature rise, and contact quality reviewed together. A cable can pass pinout and still run hot in a dense installation if the connector plating, pair resistance, or routing bundle was chosen too casually.

Which Color Code Should You Choose?

Choose the standard that matches the customer requirement or the existing installation standard. If nothing is specified, T568B is often the default in commercial practice, but T568A is equally correct when used consistently.

Use a straight-through cable for normal Ethernet links. Use a crossover cable only when the equipment requires it. Use a rollover cable only for console-style applications and mark it so clearly that nobody can confuse it with a data patch cord.

If the build involves custom length, shielding, panel transitions, or integration into a larger box build, document the complete assembly rather than only the pin order. In manufacturing, the drawing package must control connector part number, cable category, jacket material, label text, and test method, not just the colors.

Quick Selection Shortcut

If you need a short bench rule, use this:

1. Use T568B-to-T568B for most ordinary straight-through patch leads.
2. Use T568A-to-T568A when the building or customer standard requires T568A.
3. Use T568A-to-T568B only for a true crossover cable.
4. Use a reversed pinout only for a true rollover or console cable.
5. Always test for wiremap and split pairs before release.
6. For PoE or shielded industrial links, review thermal load and shield termination as part of the assembly design.

FAQ

Q: What is the standard color code for an Ethernet cable using T568B?

For T568B, pins 1 through 8 are white/orange, orange, white/green, blue, white/blue, green, white/brown, and brown. That sequence is one of the 2 recognized TIA/EIA-568 termination patterns and is commonly used for straight-through patch cords in commercial networks.

Q: Is T568A better than T568B?

No. T568A and T568B perform the same when applied consistently because the electrical pair structure is still compliant. The real requirement is that both ends of a straight-through cable match. Problems usually come from unintended mixing, not from either standard being inherently faster. On 1000BASE-T links, all 4 pairs are used regardless of whether you choose A or B.

Q: What is the difference between a crossover cable and a rollover cable?

A crossover cable swaps the orange and green pair positions by using T568A on one end and T568B on the other. A rollover cable fully reverses the pin order, so 1 maps to 8 and 2 maps to 7. Crossover supports certain legacy Ethernet device-to-device links, while rollover is typically used for console connections rather than normal data traffic.

Q: Can a cable pass continuity but still fail network testing?

Yes. A split-pair cable can show correct 1-to-1 continuity on all 8 pins and still fail because the twisted-pair geometry was broken during termination. That can cause excessive crosstalk, poor return loss, and unstable 100BASE-TX or 1000BASE-T performance. This is why production testing should include wiremap and split-pair detection, not only continuity.

Q: How much untwist is acceptable when terminating Cat5e or Cat6 cable?

As a practical manufacturing target, keep pair untwist to about 13 mm or less at the termination unless the connector system specifies a tighter limit. Excess untwist reduces pair balance and can degrade NEXT and return-loss margin, especially as performance expectations move from Cat5e toward Cat6 and Cat6A assemblies.

Q: Does PoE change how I should evaluate network cable assemblies?

Yes. Once the cable is carrying PoE, especially above 30 W, conductor resistance, contact quality, bundle heating, and connector temperature rise deserve more attention. A cable that is acceptable for low-power data can become a thermal or reliability problem in dense bundles if the copper size, plug quality, or routing environment is marginal.

If you need custom Ethernet cables, shielded industrial patch leads, or a full interconnect package that ties network I/O back to PCB assemblies and box builds, we can review the drawing, pinout, and test plan before release. Contact YourPCB for a quote or DFM review.