Does the temperature of a magnet affect its strength? Yes, for nearly every magnet used in daily life. Warming a permanent magnet weakens its pull, often permanently if it gets hot enough; cooling it generally strengthens it, but only within limits, and extreme cold can crack brittle magnet materials. The size of the effect depends on the magnet's material, its starting strength, and how close it runs to a critical temperature called the Curie point, where all magnetic order collapses.
What heat actually does inside a magnet
A permanent magnet is not a solid block of magnetism. It is a collection of microscopic magnetic regions called domains, each pointing in a preferred direction. In a strong magnet, those domains are aligned. Heat attacks that alignment from two sides.
First, thermal energy jiggles atomic magnetic moments, which are tiny compass needles tied to electron spin. As temperature rises, the jiggle competes harder with the exchange coupling that normally locks neighboring moments parallel. The material's saturation magnetization, the maximum magnetism it can produce, falls. Second, heat makes it easier for domains to rotate away from the preferred direction under an opposing field, so the coercivity, the resistance to demagnetization, drops.
The practical result is that a magnet's residual induction, the field it produces with no external load, shrinks as temperature climbs. A plain-language version: heat loosens the grip that keeps a magnet's internal compass needles pointed the same way, and the pull weakens before the magnet visibly fails.
Quantified context: how much strength moves per degree
The size of the loss is expressed as a reversible temperature coefficient, usually given per degree Celsius. For neodymium-iron-boron, the strongest common permanent magnet, the residual field coefficient is roughly minus 0.11 to minus 0.13 percent per degree Celsius near room temperature. The coercivity coefficient is worse, around minus 0.55 to minus 0.65 percent per degree Celsius.
Do the arithmetic and the sensitivity becomes concrete. A neodymium magnet loses about 1.2 percent of its surface field for every 10 degrees Celsius, or about 18 degrees Fahrenheit, of warming. Going from 68 F to 212 F, a 144 F rise or 80 C, erodes surface strength by roughly 9 to 10 percent. That is often enough to notice on a holding job and usually not enough to ruin the magnet.
Samarium-cobalt behaves differently. Its residual coefficient is smaller, near minus 0.03 to minus 0.04 percent per degree Celsius, and its coercivity holds up far better in heat. Ferrite, the cheap ceramic magnet in refrigerator magnets and speakers, sits between the two, with a residual coefficient around minus 0.2 percent per degree Celsius, which makes it noticeably more heat-sensitive than neodymium in ordinary conditions.
Magnets held in storage show another route to loss. If a magnet is stored in contact with steel, or in a stack with its own field redirected through neighboring magnets, heating it can alter the self-demagnetizing field it experiences. That can shift the operating point on its demagnetization curve and produce a permanent loss even at temperatures the material would normally survive.
The time course of a heat excursion
The sequence below is typical for a neodymium magnet rated for ordinary use, taken well past its normal operating range.
| Stage | Temperature (approximate) | What happens to magnetic strength |
|---|---|---|
| Room temperature | 68 F (20 C) | Full rated strength; reversible coefficient applies |
| Mild warming | 104 F (40 C) | Roughly 2 percent reversible loss from room baseline |
| Elevated use | 176 F (80 C) | Roughly 7 percent reversible loss; still recovers on cooling |
| Near maximum operating temperature | 212-248 F (100-120 C) | Reversible losses deepen; irreversible loss begins for lower grades |
| Above maximum operating temperature | 302 F (150 C) and higher | Permanent loss grows; domains fail to realign on cooling |
| Near Curie point | Grade-dependent, roughly 590-662 F (310-350 C) for neodymium | Magnetism collapses entirely; no recovery on cooling |
| After cooling from a severe excursion | Returns toward room temperature | Some fraction of strength is permanently gone |
The key distinction is reversible versus irreversible loss. Below a grade's maximum operating temperature, a neodymium magnet recovers most of its strength when it cools. Above that temperature, a fraction of domains have flipped and stay flipped, so the magnet returns weaker than it started. This is why a magnet that survived a hot car often still works but holds less.
Edge cases: cold, grades, and material differences
Cold is not simply the opposite of heat. Cooling a neodymium magnet increases its residual field by roughly the same coefficient, about 1.2 percent per 18 F, and also raises coercivity. That sounds like free strength until brittleness enters the picture. Neodymium magnets are mechanically brittle, and thermal shock from rapid cooling can crack or chip them. A deep freeze followed by a rapid warm-up is a realistic way to damage a magnet without exceeding any magnetic temperature limit.
Grade matters more than most buyers realize. Neodymium magnets are sold in temperature grades such as N, M, H, SH, UH, and EH. The letters encode maximum operating temperature, not base strength. An N-grade magnet might be rated near 176 F (80 C), while an EH-grade of the same size and shape can survive well past 392 F (200 C). The higher grades achieve that by adding dysprosium or terbium, which raises coercivity but can slightly reduce the achievable residual field. Two magnets that look identical can therefore behave very differently in an engine bay.
Ferrite magnets show the sharpest practical cliff. Their maximum operating temperature is often around 482 F (250 C), but their Curie point is only around 842 F (450 C), so a ferrite magnet that gets seriously hot loses a large share of its pull permanently. Alnico magnets, used in guitar pickups, are the reverse edge case: they tolerate high temperatures but demagnetize easily under an opposing field, so a hot alnico magnet near a strong opposing magnet can lose strength it will not fully regain.
The counterintuitive detail is that a magnet can lose strength in heat even when its own field is doing the damage. A magnet whose shape produces a large self-demagnetizing field, such as a thin disc with widely separated poles, sits closer to the edge of its demagnetization curve. Warming that magnet lowers its coercivity until its own internal field flips domains. The same material in a long, closed shape survives the same temperature intact.
Misconceptions worth retiring
The most common error is treating temperature ratings as a single number that applies to every magnet of a given material. The maximum operating temperature is a property of a specific grade and shape, not of "neodymium" in general. A cheap online listing that says only "neodymium" tells you nothing about heat tolerance.
The second error is assuming heat damage is always visible. It is not. A magnet that has lost 15 percent of its strength through overheating looks, sounds, and feels exactly like a good one. Only a pull test or a gaussmeter reading reveals the loss.
The third is believing cold cooling is always safe. It strengthens standard permanent magnets, but it also makes brittle materials more prone to fracture, and it does not help every technology at all.
Magnets whose strength is produced by an electric current, electromagnets, follow a completely different rule. Their field comes from current flow, and heating raises the wire's electrical resistance. At a fixed supply voltage, hotter copper carries less current, so the field weakens with heat for a reason that has nothing to do with domains. Temperature and magnet strength are linked in both cases, but the mechanism differs.
Red flag: buying an unrated "high-strength" neodymium magnet for a high-temperature application such as an exhaust system or motor housing. Red flag: assuming a magnet that survived a hot car still meets its original pull rating. Red flag: rapid cooling of a brittle magnet in an industrial freezer or with dry ice. Red flag: using a ferrite magnet near a heat source above roughly 482 F (250 C). Red flag: rejecting a magnet as defective after one hot cycle without a pull test.
Practical actions for anyone working with magnets
Measure before you trust. A gaussmeter or a simple pull test with a known mass and a spring scale gives a baseline. Recheck after any thermal event. A loss of more than a few percent below the rated maximum operating temperature is a sign the grade was misjudged.
Match the grade to the hottest condition, not the average. Add margin: choose a maximum operating temperature at least 86 F (30 C) above the hottest expected surface. For a magnet that sees 176 F (80 C), an H-grade at 248 F (120 C) leaves real headroom.
Keep heat sources away from the poles. A magnet's own field does not heat it, but nearby solder joints, motors, and lamps do. Shield brittle magnets from direct radiant heat and avoid placing them in enclosed spaces that build heat.
Store stacked with steel keepers. A keeper closes the magnetic circuit and lowers the self-demagnetizing field, which protects strength during storage and transport and makes heat excursions less damaging.
When cold is the goal, cool slowly. If a magnet must be cooled, bring it down gradually and warm it gradually. Fast thermal cycling is the main cause of cracking in brittle grades.
For electromagnets, derate the current as temperature rises. Copper resistance climbs with temperature, so a coil that produces its rated field cold will produce less when hot at the same supply voltage. Check the coil's rated temperature rise and size the supply accordingly.
Limits of what can be concluded
The coefficients cited here are typical ranges for commercial grades, not universal constants. Real magnets vary with composition, processing, shape, and how they are measured. A gaussmeter reading at the pole face is not the same quantity as the material's residual induction, so measured field ratios can differ from theoretical coefficients.
The maximum operating temperatures quoted by manufacturers are application-dependent. They assume a certain load line set by the magnet's shape and the circuit it sits in. A disc magnet used alone and the same magnet in a closed steel circuit do not share the same safe temperature, even with identical chemistry.
There is also no reliable way to infer a magnet's thermal history from its surface field after the fact, unless the original field was recorded. Without a baseline, a magnet that has been overheated looks exactly like a weaker grade that never got hot. The only safe approach is to record strength when the magnet is new, protect it from heat and thermal shock, and recheck after any excursion rather than assuming the rating told the whole story.