Quick Answer / Summary: 300 ft equals 91.4 m and fits the 100 m channel only if the permanent link stays at 90 m or less. Two 5 m patch cords plus the horizontal run must total 100 m, not 100 m each. For 802.3at at 30 W, use 23 AWG conductors, since a 91.4 m run of 24 AWG adds noticeably more loop resistance and drops volts the load never sees. Choose Cat6a when 10GBASE-T shares the route.
Three hundred feet is 91.4 m, which leaves 8.6 m of the 100 m channel limit for everything that is not the horizontal run: patch cords, a consolidation point and the service loop at the outlet. That margin disappears quickly on a real installation. This article sets out the channel budget for a 300 ft run, the voltage drop an 802.3at load sees at that length, and the conductor gauge that keeps the link inside specification.
What 300 ft means against the 100 m limit
The 100 m limit applies to the whole channel: horizontal cable from the patch panel to the outlet, plus the patch cord at each end. A 91.4 m horizontal cable with two 5 m cords totals 101.4 m and is out of specification before any tester is connected. Copper resistance and crosstalk both rise with length, so the limit is not a marketing number.
Channel versus permanent link
The permanent link is the installed cable with its terminations, tested without patch cords; the channel includes them. A tester reports both. For a 300 ft run, the permanent link must test under 90 m so the channel still has room for cords at both ends.
Where the 8.6 m actually goes
Two 5 m patch cords take 10 m, which already exceeds the 8.6 m margin. That is why the practical answer is a horizontal run of 90 m or less: the cords are part of the budget, and a service loop at the outlet counts against the same number.
The channel budget line by line
The table shows how a 91.4 m cable plus ordinary cord lengths lands against the limit. Every row is a physical length that a tester will measure or that a designer should allow for.
| Element | Length | Count | Running total |
|---|
| Horizontal cable, 300 ft | 91.4 m | 1 | 91.4 m |
| Patch cord at the panel | 3.0 m | 1 | 94.4 m |
| Patch cord at the desk | 3.0 m | 1 | 97.4 m |
| Service loop at the outlet | 1.5 m | 1 | 98.9 m |
| Remaining margin | 1.1 m | 0 | 100.0 m |
PoE over 91 m: voltage drop and conductor gauge
23 AWG copper is about 0.066 ohm per metre per conductor, so a 91.4 m run adds roughly 12 ohm of loop resistance. At the 0.6 A that 802.3at draws for a 30 W load at 50 V nominal, that resistance removes several volts before the powered device sees them. A 24 AWG cord at the same length adds more, and the total is what the device must tolerate.
When 23 AWG is mandatory
Treat 23 AWG as the default for any 802.3at or 802.3bt run beyond 60 m, for bundled cables where several pairs carry current inside one jacket, and for runs in insulated walls where heat cannot escape. Bundled 24 AWG cable at 30 W per port is where installations fail.
When Cat6a is the safer buy
Cat6a buys two things here: 500 MHz of rated bandwidth for 10GBASE-T at the full 100 m, and heavier conductors in most constructions. If the 300 ft run may later carry 10G, or passes near motors and dimmers, Cat6a removes the argument during installation.
Failure modes at this length
Most 300 ft failures are termination and routing faults rather than cable faults. Untwisted pairs at the jack, a cord stapled tight around a corner, or a bundle run alongside a power circuit all cost margin that the length has already spent.
Crosstalk and alien crosstalk
Pair-to-pair crosstalk is controlled by the twist and the jack; alien crosstalk comes from neighbouring cables in the same bundle. At 91.4 m the signal has lost more of its amplitude, so the same noise floor hurts more than it would at 30 m.
The field test that settles it
Ask for a channel test that reports insertion loss, pair-to-pair and power sum crosstalk, and return loss at 250 MHz for Cat6 or 500 MHz for Cat6a. A pass at those frequencies on the channel is the evidence a warranty claim will need.
Two installations at 300 ft
Both jobs sat close enough to the limit that the last few metres of cord decided the outcome.
Warehouse access point on a 30 W port
An access point at the far end of a warehouse was fed by a 91 m Cat6 run with 24 AWG patch cords at both ends. It booted but dropped its radio at peak load. Re-terminating into 23 AWG solid cable and shortening the panel cord to 2 m held the port at 30 W.
Home office to a detached garage
A 300 ft Cat6 run from a house to a garage switch linked at 1000BASE-T but dropped packets under load. The cable had been pulled tight around two corners. Re-routing with a 25 mm bend radius and re-terminating both ends produced a clean channel test at 250 MHz.
Frequently Asked Questions
Is a 300 ft Ethernet cable too long?
No, if it is the only cable in the channel. 300 ft is 91.4 m and the limit is 100 m, so two patch cords of 3 m each still fit. Anything past 300 ft plus the cords exceeds the limit.
Does a 300 ft cable reduce speed?
Not below its category rating when the channel tests clean. Speed drops when length pushes insertion loss or crosstalk past the limit, and the link then retrains to 1000BASE-T or 100BASE-TX.
Can I run 30 W PoE over 300 ft?
Yes, with 23 AWG conductors and short patch cords. Check the powered device input range, because a 91.4 m run removes several volts, and bundled cables run warmer than a single cable.
Should I buy Cat6 or Cat6a for 300 ft?
Cat6 handles 1000BASE-T over 100 m and 10GBASE-T only to 55 m. If 10G is possible on that route, Cat6a is the correct choice because it holds 500 MHz across the full 100 m.
How do I test a 300 ft run?
Use a channel tester at the rated frequency, 250 MHz for Cat6 or 500 MHz for Cat6a, and keep the report. Margins near zero for insertion loss mean the run sits on its limit.
Related resources: how a 75 ft run changes the channel budget; shorter 30 ft runs and where they sit in the limit; cabling for access points with PoE budgets
