IP66, IP67 and IK10: What a Camera's Weather Rating Actually Promises

Pick up any outdoor camera spec sheet and you will find an IP rating near the top, presented as reassurance. It is reassurance of a very specific and quite narrow kind, and the gap between what buyers think it promises and what it actually certifies is where a surprising number of dead cameras come from. The short version: IP ratings describe how a sealed enclosure performed in a defined laboratory test. They say nothing at all about British weather over five years. Last updated: November 2026. The two digits, and what each one is IP comes from IEC 60529, and the two digits are independent tests. The first is solids, the second is liquids. First digit (dust and solids). On any serious outdoor camera this will be a 5 or a 6. A 5 means dust protected: some dust may get in, but not enough to interfere with operation. A 6 means dust tight, which is what you want on a dusty yard, a grain store, a quarry or a site with constant vehicle movement. Second digit (water). This is the one people misread, because it is a ladder of different tests rather than a single scale of wetness: 4 — splashing water from any direction. 5 — water jets from a 6.3mm nozzle. 6 — powerful water jets from a 12.5mm nozzle. 7 — temporary immersion, one metre deep, thirty minutes. 8 — continuous immersion at a depth the manufacturer states. Why IP67 is not automatically better than IP66 This is the genuinely useful thing to know, and it catches out people who assume a bigger number is strictly superior. The tests are separate. A device certified IP67 has passed the immersion test. It has not necessarily passed the powerful-jet test, because that is a different test and immersion does not include it. So a camera rated IP67 and nothing else has no certified claim against a pressure washer, while an IP66 camera does. That is why you will frequently see cameras listed as IP66/IP67, carrying both. Both tests were run, both passed, and the device is covered for jets and for brief immersion. If a spec sheet quotes only one of them, it is worth knowing which risk your site actually presents. Practically: a camera on a wash-down bay, a livestock building or a yard that gets jetted needs the 66. A camera low on a post where standing water can rise over it, or in a position that floods, wants the 67. Plenty of sites want both, and that is a perfectly reasonable thing to specify. IK is a different rating entirely, and it is about impact IP says nothing about someone hitting the camera. That is IK, from IEC 62262, and it is measured in joules of impact energy: IK07 — 2 joules. IK08 — 5 joules. IK09 — 10 joules. IK10 — 20 joules, the top of the scale. The rule of thumb is simple: if a member of the public can reach the camera, specify IK impact resistance; if they cannot, spend the money elsewhere. A camera at four metres on a pole in a private yard does not need IK10. A camera in a school corridor, a car park stairwell, a bus shelter or a community building porch very much does, and our school and education site guide goes into where vandal-resistant housings earn their cost. Mounting height is the cheapest vandal resistance available. Use it first. What the ratings do not cover at all Here is the part the spec sheet does not tell you, and it accounts for most real-world failures on UK sites. Condensation. IP testing is about water getting in from outside. It says nothing about moisture already inside the housing condensing on the lens when the temperature swings, which is the classic British autumn failure: a camera that looks fine at midday and shows a milky fog every morning. The defences are a properly sealed enclosure, an intact desiccant pack, and not opening the housing on a damp day and then closing it again on the damp air you just let in. UV and plastic ageing. Years of sunlight make unprotected plastics chalky and brittle, and brittle housings crack at the fixing points. This is a materials question, not an IP question. Temperature range. A separate specification, and one that matters more for batteries than for the camera. Check the stated operating range if the camera sits somewhere that genuinely freezes, and see CCTV power planning for how cold affects an off-grid install. Salt and corrosion. Coastal and marine sites eat fixings, brackets and screw heads long before they trouble the camera body. Stainless fixings are a small cost and a large difference. Corrosive and explosive atmospheres. Entirely outside IP's scope. If the site has a zone classification, you need certified equipment, which is the subject of our ATEX CCTV specification guide. No IP rating substitutes for it. The cable gland is where it actually leaks If you remember one practical thing from this, make it this one. The housing is rarely the failure point. The cable entry is. An IP66 camera on the end of a badly dressed cable is an IP66 camera with a hole in it. The things that go wrong are consistent and avoidable: A gland tightened onto the wrong cable diameter, so it grips nothing and seals nothing. An unused gland left open, which is simply a vent into the housing. Connectors joined in open air rather than inside a rated junction box. This is the single most common one. No drip loop. Cable entering a housing from above lets water track down the sheath and straight in. Loop the cable below the entry so gravity does the work. Self-amalgamating tape skipped on outdoor connector joints. Our camera siting and mounting guide covers the mounting discipline in full, and it is worth more to the lifespan of an install than two points of IP rating. While you are up the pole, the same logic applies to antennas: a water-filled antenna connector is a slow signal failure, which is covered in why 4G cameras drop offline. Specifying it, in one pass Dust tight as standard. Ask for a 6 as the first digit on anything outdoors. Match the second digit to the real risk. Jetting and driving rain means 6. Standing water and flood risk means 7. Both is common and sensible.…