The question usually arrives at the worst possible moment. The cameras are chosen, the recorder is on order, the walls are open — and someone asks whether the cable is the right one. The instinct is to buy the highest category available and stop worrying. That instinct costs money and, on its own, does not solve the problem it was meant to solve, because the category on the jacket is not what limits most camera installations.
Part of the confusion comes from how power and data now share the same cable. Power over Ethernet is not a separate wire; the standard that defines multi-pair PoE describes raising power delivery "by utilizing all four pairs in the specified structured wiring plant" — meaning the same twisted pairs that carry the video stream also carry the current (IEEE 802.3bt-2018). That single fact is why cable choice matters for surveillance in a way it does not for a desktop PC. A bad patch cord on a laptop costs you speed. A bad permanent link on a camera costs you an image at 2 a.m.
So the question worth answering is narrower than "which cable is better": for a 4K PoE camera, where does the cable category actually change the outcome, and where is it irrelevant? This article works through the bandwidth arithmetic with real bitrate figures, isolates the three situations where Cat6 genuinely earns its price, and explains why the conductor inside the jacket matters more than the category printed on it.
The Short Answer, in One Paragraph
For a 4K PoE camera, Cat5e is enough. A 4K stream at its maximum 8 Mbps uses well under one per cent of what Cat5e carries, and even sixteen cameras at full bitrate stay inside a quarter of it. Buy Cat6 when your longest run approaches the distance limit, when the cable shares a pathway with electrical power, or when you expect multi-gigabit networking later. Buy solid copper either way, and never buy CCA.
What Each Cable Category Supports
The categories are defined by a rated bandwidth and a guaranteed data rate at a stated distance. Almost every argument about camera cable is really an argument about the second column.
| Category | Rated bandwidth | Data rate and the distance it holds | Typical conductor |
|---|---|---|---|
| Cat5e | 100 MHz | 1 Gbps to 100 m (328 ft) | 24 AWG |
| Cat6 | 250 MHz | 1 Gbps to 100 m; 10 Gbps to roughly 55 m | 23 AWG |
| Cat6a | 500 MHz | 10 Gbps to the full 100 m | 23 AWG |
Two intermediate rates are also standardized over this cabling and are worth knowing about: 2.5GBASE-T and 5GBASE-T, which run over Cat5e and Cat6 respectively to the full 100 m. They matter for surveillance because a recorder with a 2.5 GbE uplink can aggregate a large camera count without Cat6a.
For a camera system, the practical reading is this: at 1 Gbps, Cat5e and Cat6 are identical. The difference only appears when you try to push 10 Gbps through the cable, and then the distance collapses to roughly half of the 100 m norm.
Why the common distance figures disagree
Search this topic and you will find Cat6 quoted as good for 10 Gbps at 55 m, and quoted elsewhere as good for 10 Gbps at only 33.5 m or 50 m. Both figures circulate because they describe different test conditions, and no competitor page bothers to reconcile them.
The shorter numbers reflect the basic permanent-link channel without special alien-crosstalk mitigation. The 55 m figure assumes a favourable installation where crosstalk between adjacent cables is controlled. If you are designing to the standard rather than to the marketing, plan on the shorter number and treat anything better as margin. Either way, the figure is measured in tens of metres, not hundreds — which is why the next section matters more than this one.
What the 100 metre rule actually includes
The 100 m channel limit is not 100 m of cable hanging on a hook. It is the total channel: up to 90 m of permanent link in the wall plus up to 10 m of patch cords at either end. Installers who measure 100 m of cable and then add two patch cords have already exceeded the channel they were designing to. Measure the permanent link, then count the patch cords against the allowance.
The Arithmetic Behind a 4K Camera on Cat5e
This is the calculation competitors skip. Nearly every page says "4K is about 30 Mbps, and Cat5e does 1,000 Mbps, so you are fine." That comparison is directionally right and numerically loose, which makes it easy to distrust.
Single camera, two codecs
Start with what a camera actually produces. A typical 4K camera's own datasheet caps its stream at 8,192 kbps — about 8 Mbps — and vendor planning figures put a 4K camera at roughly 4 to 6 Mbps under H.265 and 8 to 12 Mbps under H.264, with some configurations reaching higher.
Against a 1 Gbps link, 8 Mbps is 0.8 percent. Even taking a pessimistic 16 Mbps, a single 4K camera occupies 1.6 percent of what Cat5e carries. The cable is not remotely close to being the constraint.
Scaling to eight or sixteen cameras
Single-camera math is where most pages stop, and it is the wrong unit of analysis once you have a recorder. Cameras aggregate.
| Camera count | Bitrate per camera | Aggregate | Share of a 1 Gbps link |
|---|---|---|---|
| 4 | 8 Mbps | 32 Mbps | 3.2% |
| 8 | 8 Mbps | 64 Mbps | 6.4% |
| 8 | 16 Mbps | 128 Mbps | 12.8% |
| 16 | 8 Mbps | 128 Mbps | 12.8% |
| 16 | 16 Mbps | 256 Mbps | 25.6% |
Even the heaviest row — sixteen 4K cameras all streaming at the upper end of their range — leaves three quarters of the link unused. Run half the cameras on H.265 at 5 Mbps and the real figure lands closer to 8 percent.
Note what this table is really measuring: the aggregate figure is the load on the recorder's uplink, not the load on any single cable. Each cable carries one camera. That distinction is the whole reason a 100 m run of Cat5e is not the bottleneck most buyers imagine.
Where the recorder becomes the bottleneck, not the cable
If something does run out of headroom, it is almost always the recorder's ingest capacity rather than the cable. Recorders publish incoming bandwidth ceilings, and an eight-channel unit capped at 80 Mbps will not accept eight full-rate 4K streams at high bitrate — the recorder, not the cable, forces the stream settings down. That is a specification worth checking before you blame the wiring.
Where Cat6 Earns Its Premium
Cat6 costs more, and for many installations that money is better spent on the hard drive. Three situations genuinely change the calculation.
Runs that sit close to the distance limit
Under 100 m, the cable category is nearly irrelevant to a camera. As a run approaches the limit, the extra headroom and tighter crosstalk performance of a 23 AWG Cat6 cable stop being theoretical. The failure mode at the edge of the limit is not a dead camera but an intermittent one — a stream that drops frames in wind or rain, which is maddening to diagnose because it tests fine on a warm afternoon.
Paths that share space with power
Unshielded Cat6 gains over Cat5e from a plastic spline separating the pairs and from tighter twist rates, which improves resistance to crosstalk between cables. It does not become magically immune to electromagnetic interference. If a data run is going to share a pathway with mains wiring, motors, or variable-frequency drives, the correct answer is usually shielded cable — F/UTP or S/FTP — installed with proper grounding, not simply a higher category.
Installations heading toward multi-gigabit
If the same cable will later carry a 10 Gbps backbone, or the recorder will be upgraded to a 10 GbE uplink, Cat6 buys you the option — within its distance limit. Cat6a buys it to the full 100 m. For a camera feed today this is future-proofing, not a requirement; be honest about whether the upgrade is actually on your roadmap.
When Cat6 is simply the wrong purchase
If every run is under 50 m, the camera count is modest, and the recorder's uplink is 1 GbE, Cat6 changes nothing except the invoice. The money is better spent on a larger drive, which extends how far back your footage reaches — a benefit you will notice every week, unlike headroom you never touch.
Copper-Clad Aluminum and What It Does to PoE
This is the part of the decision that actually causes failed installations, and it is the least covered in every article on the first page of results.
What CCA actually is
Copper-clad aluminum is a conductor with an aluminum core and a thin copper skin. It looks identical to copper cable in a photo, weighs less, costs far less, and is sold through the same channels with the same category markings on the box. The savings are real; so is the reason it is cheaper.
What higher resistance does to a camera at night
Copper-clad aluminum has substantially higher DC resistance than copper of the same diameter — the widely cited figure is around 55 percent higher. Resistance is exactly what you do not want in a cable that is also carrying power, because current through resistance produces a voltage drop, and the drop grows with both distance and current draw.
That is how CCA fails a camera system in practice. On a bench it works. At 4 a.m., when the infrared illuminators switch on and the camera's current demand rises, the voltage arriving at the device sags below what it needs and the camera reboots, drops off the network, or produces a black frame. The symptom looks like a camera fault, so the camera gets returned and the cable — the actual cause — stays in the wall.
Testing labs measure this directly. A Fluke Networks application note documents a CCA link measured at 31.08 ohms against a 21.00 ohm limit over a 295 ft run — a clear failure — and states plainly that these cables "do not support PoE applications due to their increased DC Resistance" (Fluke Networks, CCA wire application note).
Why CCA cannot legally carry a category marking
A cable sold as Cat5e or Cat6 is supposed to meet a standard that requires solid copper conductors. The relevant documents — the National Electrical Code, UL 444, the TIA-568 series, and ISO/IEC 11801 — specify copper for multi-conductor communications cable. Because of that, a CCA cable cannot hold a valid safety listing for the spaces where permanent cabling typically goes.
The practical consequence, per the same source: installation of multi-conductor communications cable made with CCA conductors behind walls or in enclosed spaces "is likely to be a code violation in every jurisdiction in the country." If you are comparing a suspiciously cheap box against a normal-priced one, the price gap is often the whole answer.
Heat, Bundling, and How Many Cables You Can Group
PoE puts current through cables that, in a surveillance install, are frequently bundled together in a raceway or pulled through a single hole. Current heats the conductor, and heat in a bundle has nowhere to go.
How the standard handles it
The PoE standards acknowledge this by derating power at higher ambient temperatures: roughly 5 °C of derating for two-pair Type 2 operation and about 10 °C for the four-pair Type 3 and Type 4 modes. The bundling guidance in the TIA's TSB-184 goes further, and the numbers are more dramatic than most installers expect. A Cat5e or Cat6 unshielded bundle that is fine at a 45 °C ambient is limited to a small fraction of that cable count at 60 °C — in the published table, twenty-four cables becomes six. Shielded Cat6a bundles fare better, but they still derate.
What that means for a camera installation
Conduit in a sun-exposed attic, or a raceway above a ceiling, can easily sit in the range where the cable count matters. The practical guidance is unglamorous: split large bundles into separate pathways where you can, avoid filling conduit completely, keep runs away from hot equipment, and treat the bundle itself as a design constraint rather than an afterthought.
Jacket Ratings and Where the Code Requires Them
The category tells you what the cable can carry. The jacket rating tells you where you are allowed to put it. Buyers rarely check this, and it is the second most common reason an inspection fails.
The ratings and the spaces each one allows
| Marking | Rating | Where it is permitted |
|---|---|---|
| CMP | Plenum | Air-handling spaces, including above drop ceilings. Tested to NFPA 262 |
| CMR | Riser | Vertical shafts between floors. Tested to UL 1666 |
| CM | General purpose | Ordinary occupied spaces such as offices |
| CMX | Limited use | Restricted exterior and perimeter applications |
All four sit under ANSI/UL 444, the safety standard for communications cable. The compliance requirements are specific: plenum cable test criteria under NFPA 262 include a maximum peak optical density of 0.50, an average optical density of 0.15, and flame spread of five feet or less; riser cable must not propagate flame to 12 feet under UL 1666.
Which rating a camera install actually needs
For most residential installs, the run goes through wall cavities and an attic, which lands on CM or CMR depending on local rules and whether the space is used for air handling. If the cable passes above a suspended ceiling that returns air to the HVAC system, that is a plenum space and you need CMP — this is the one that catches people out, because the ceiling looks like any other ceiling.
Outdoors is its own category. Direct sunlight degrades a standard PVC jacket, so exposed exterior runs want a UV-resistant outdoor or CMX-rated jacket, and anything buried needs a jacket rated for direct burial rather than a standard one in a piece of conduit.
The Buying Specification: Reading the Jacket Print
Everything you need to know is printed on the cable itself, in a string of abbreviations that most people skip past. Once you can read it, the marketing claims stop mattering.
Reading the print string, part by part
A conforming print runs along the jacket in roughly this order, and this is an example of the format rather than a specific product: CMR U/UTP CAT6 23AWG 4PR 75C 100% OFC 250MHz. Read it in six parts.
The first token is the fire rating — CMR here, meaning riser. The second describes the construction: U/UTP means unshielded twisted pair, while F/UTP or S/FTP would indicate shielding. The third is the category. The fourth is the conductor gauge, where 23 AWG is typical of Cat6 and 24 AWG of Cat5e. The fifth is the pair count — four pairs, which is what you need for four-pair PoE in the higher power modes. The sixth tells you the conductor material, and this is the token that matters most: OFC means oxygen-free copper, and you want copper stated explicitly rather than left vague. The final figure is the rated bandwidth.
The one token you should be suspicious of is a missing conductor description, or the presence of the letters CCA, which describe exactly the conductor you do not want in a PoE run.
The faults that make a cable unsuitable regardless of category
Four things disqualify a cable no matter what the box claims. A stranded-conductor patch cable used as a permanent in-wall run — patch cords are meant to be short and flexible, not installed. Copper-clad aluminum in any PoE application. A missing or unreadable print string. And a jacket rating that does not match the space it is going into.
Matching the Cable to the Camera Plan
Specifications become a decision when there is a building attached to them. What follows is a configuration for each situation rather than a product list, because the right cable is defined by numbers you can check.
Four to six cameras on a single-family home
Solid-copper Cat5e, unshielded, 24 or 23 AWG, CM or CMR depending on the pathway. Every run under 100 m, no shared pathways with mains wiring, and no ambition to run 10 Gbps. This is the majority of residential installs, and Cat6 adds cost without changing an outcome. Buying solid-copper Cat5e rather than a cheaper CCA box is the decision that matters here.
Eight cameras with mixed resolutions
Cat5e remains sufficient on bandwidth grounds. If the plan mixes 4K cameras with a couple of lower-resolution units, aggregate demand is still a small fraction of the link. Spend the category premium on storage instead — a larger drive extends your retention window, which is the constraint you will actually run into.
Runs that approach the distance limit
Once a run stretches toward the upper end of the 100 m channel, step to Cat6 for the extra headroom, and measure the channel honestly — permanent link plus patch cords. If the distance is unavoidable and genuinely long, a mid-run switch that re-times the signal, or fibre for the backbone, solves it more reliably than any copper category.
Outdoor paths, sunlight, and direct burial
Specify a UV-resistant outdoor or CMX-rated jacket for anything exposed, and a direct-burial rated jacket for anything underground. Standard indoor PVC jacket material degrades in sunlight, and a cracked jacket on an exterior run lets water into the pair and takes the camera down.
Paths that share space with heavy electrical loads
Where a camera cable must share a pathway with mains power, motors, or drives, use shielded cable — F/UTP or S/FTP — and ground the shield properly at one end. A higher unshielded category is not a substitute for shielding, and an improperly grounded shield can make interference worse rather than better.
One last note on sourcing: PoE kits increasingly include or offer the cable as an add-on — 4COVR lists security camera cable alongside its PoE systems, for example — and that is a convenient way to get matched lengths. Whichever route you take, the category and conductor printed on that jacket is what determines whether a long run still powers a 4K camera after dark.
Questions People Ask Before They Buy
Can a 4K camera run on Cat5e?
Yes, comfortably. A 4K camera's stream sits in the range of 4 to 12 Mbps depending on codec and settings, against 1,000 Mbps of capacity on Cat5e. The cable is not the limiting factor for 4K. Distance, conductor quality, and the recorder's ingest ceiling are.
Does Cat6 carry more power than Cat5e?
Not in a way that affects a camera. PoE power levels are set by the standard, not by the cable category — 802.3af delivers up to 15.4 W at the port, 802.3at up to 30 W, and the 4-pair 802.3bt modes 60 W and 90 W. What Cat6 does help with is heat and voltage drop, because its thicker 23 AWG conductors have less resistance over distance. That is a real benefit on long runs, and irrelevant on short ones.
What happens if I run 4K cameras on cheap CCA cable?
It often works at first, then fails intermittently. Higher resistance means a larger voltage drop, which shows up when the camera's current demand rises — typically at night when infrared illuminators engage. The camera browns out, reboots, or drops offline. Because the camera looks like the faulty component, the cable is rarely suspected.
Do I need shielded cable near electrical lines?
If a run must share a pathway with power, motors, or variable-frequency drives, shielded cable with proper grounding is the right answer. Keeping data and power in physically separate pathways is better still where the layout allows it.
How do I tell whether a cable is really copper?
Check the print on the jacket and the weight of the box. A 1,000 ft box of solid-copper Cat5e currently runs roughly $139 to $170, and Cat6 solid copper roughly $185 to $230. Boxes selling far below those ranges are frequently copper-clad aluminum, and the print string will either say CCA or omit the conductor description entirely.
Is it worth replacing existing Cat5e with Cat6?
Almost never for cameras. If your existing Cat5e is solid copper, correctly terminated, and within distance limits, it will carry 4K camera traffic with enormous margin. Replace it if it is CCA, damaged, or if you are moving to a multi-gigabit backbone for other reasons.
Where This Leaves the Decision
The category on the jacket is the least interesting variable in this decision, and it is the one everybody argues about. For 4K cameras, Cat5e carries the data with room to spare — the arithmetic is not close.
What actually decides whether an installation works is the conductor and the pathway. Solid copper rather than copper-clad aluminum, because the cable is carrying power as well as data and resistance is what breaks a camera at night. A jacket rating that matches the space the cable runs through, because that is what an inspection checks. Sensible bundling and separation from power, because heat and interference are the two failure modes that a spec sheet will never show you. And honest measurement of the channel, because 100 m is the whole path, not the length of the run.
Buy Cat6 when your longest run pushes toward the distance limit, when the cable will later carry a multi-gigabit backbone, or when shielding and thicker conductors are solving a problem you actually have. Otherwise buy good Cat5e, spend the difference on storage, and check the print on the jacket before it goes in the wall.