Best Thermometer for Candle Making: Probe in the Wax, Infrared on the Jar

"An infrared thermometer is a surface temperature tool – period." That's not a skeptic talking. It's ThermoWorks (opens in new tab), a company that sells a whole shelf of infrared guns, in its own myth-busting article. The same piece gives an example every candle maker should read twice: point a fixed-emissivity infrared thermometer at the side of a stainless steel pot of boiling water, and "you might get a reading closer to 100°F (38°C) than 212°F (100°C)."

Now picture the classic candle bench. An aluminum pour pitcher sits in a double boiler, and an infrared gun is aimed at it because it's quick and the laser dot looks precise. On ThermoWorks' emissivity table (opens in new tab), polished aluminum rates 0.05, which means it reflects nearly all the infrared that hits it. That gun isn't reading the wax. It's mostly reading the room.

So here's the short answer. The best thermometer for candle making is a digital instant-read probe that goes in the wax and publishes its accuracy, ideally ±1°F or better between about 100°F and 200°F. An infrared gun is an optional second tool, and its job is checking jars. The winner isn't the priciest or the most exotic option. It's the one that publishes a number you can check.

What a candle thermometer actually has to measure

Before comparing tools, pin down the job. Candle wax has three temperatures that matter, and your wax supplier already wrote them down for you.

The fragrance-add temperature. CandleScience gives the same instruction on every wax page in the table below: heat to 185°F (85°C), add fragrance, remove from heat, stir for two minutes. When a Golden Brands 464 candle has a weak scent throw, the troubleshooting list on its Golden Brands 464 page (opens in new tab) asks you to check that fragrance really went in at 185°F.

The pour temperature. This is where waxes differ, and not by a little:

The adjustment step. This is the number that decides which thermometer to buy. When a GB 464 candle comes out with a rough top, CandleScience suggests pouring 5 to 10°F hotter. For IGI 4630 jars with poor glass adhesion, it suggests experimenting with pour temperatures of plus or minus 5°F. So the fixes CandleScience hands you come in 5°F steps.

Meet Dani, a composite example: a soy candle maker who pours GB 464 in a spare bedroom. Her candle tops are coming out rough, so she tries the supplier's fix: this week she'll pour at 140°F instead of 135°F. Suppose her thermometer can be 4°F off in either direction. Then her "140°F" pour could have gone in anywhere from 136°F to 144°F, and last week's "135°F" anywhere from 131°F to 139°F. The two ranges overlap. Her real change could be anything from 3°F cooler to 13°F hotter. The test might have gone the wrong direction, or overshot by more than double.

That 4°F isn't hypothetical. USDA's Food Safety and Inspection Service says most food thermometers (opens in new tab) "will give an accurate reading within 2 to 4 degrees F." For roasting a chicken, that's fine. For a 5°F wax experiment, it can swallow most of the experiment.

A useful rule of thumb: buy a thermometer whose published error is no more than about a fifth of the smallest adjustment you plan to make, so the thermometer's error can't account for more than a small slice of the change. For 5°F steps, that means ±1°F or better, in writing.

Five candle thermometers, weighed

Candle suppliers commonly sell one or two of these, and it's easy to end up owning three. Here they are side by side, using the figures cited in the sections below, followed by how each one does on the actual job.

Glass clip thermometer

The traditional choice, and CandleScience's soy candle making kit (opens in new tab) includes a glass thermometer. CandleScience sells an 8-inch glass thermometer (opens in new tab) rated 100°F to 400°F with an adjustable pan clip, the same range FSIS describes for candy and deep-fry thermometers.

Pros:

• Clips to the pot and stays put, so you can watch wax climb toward 185°F without holding anything
• No battery to die halfway through a pour
• The 100°F to 400°F range covers every wax in the table above

Cons:

• Hard to read to the degree. You're judging a colored line against small markings, often through steam
• FSIS says liquid-in-glass thermometers should sit at least 2 inches deep. A small test batch in a pour pitcher might not be that deep
• The product page publishes a range but no accuracy figure
• It's glass, sitting next to hot wax

Best for: watching the melt in a double boiler. It's a poor fit for deciding a pour to within a few degrees.

Dial (bimetal) thermometer

The pocket thermometer with a round face and a metal stem. It works through a coil of two bonded metals, and according to FSIS that coil senses temperature "from its tip and up the stem for 2 to 2 1/2 inches." The reading is an average along that whole length.

Pros:

• Cheap and tough
• FSIS notes many models have a calibration nut under the dial, so you can correct them

Cons:

• FSIS gives the instant-read version 15 to 20 seconds to reach a reading
• The averaging is a problem in shallow wax. If the batch is shallower than the sensing coil, part of the "wax" reading is the air above it
• ThermoWorks says dial thermometers should be calibrated "very regularly if not daily," (opens in new tab) compared with about once a year for many digital ones

Best for: very little at the candle bench. Every job it can do, a digital probe does faster and reads more clearly.

Digital instant-read probe (thermistor)

A thin steel probe with a digital display. This is the category most candle makers should buy from, and CandleScience, for one, sells a digital probe alongside its glass thermometer. CandleScience's own digital thermometer (opens in new tab) has a 5.6-inch probe that, in the listing's words, measures "the internal temperature of liquids—not just the surface temperature as with infrared thermometers."

What separates good probes from fine ones is whether the maker publishes accuracy for your temperature range. The CandleScience listing gives a range of -58°F to 572°F and no accuracy figure. ThermoWorks' ThermoPop 2 (opens in new tab) publishes a table: ±1°F from 14°F to 208°F, loosening to ±4°F from 392°F to 572°F, with a 2 to 3 second reading on the 4.5-inch probe and a NIST-traceable calibration certificate in the box. Every candle temperature in the table above falls inside the ±1°F band.

Pros:

• Reads the wax itself, not the surface
• A few seconds per reading, and a number instead of a line to squint at
• The better models publish accuracy by range, so you can check it against the 5°F rule

Cons:

• Needs a battery
• Accuracy varies between brands, and some, including CandleScience's own probe, don't publish it
• You have to hold it or clip it; most aren't made to sit in a pot the whole time

Best for: almost everyone. This is the thermometer that should decide when fragrance goes in and when you pour.

Thermocouple probe

The fastest contact thermometer you can buy. FSIS says thermocouples "reach and display the final temperature the fastest — within 2 to 5 seconds," and ThermoWorks' Thermapen ONE (opens in new tab) claims full readings in one second or less, with ±0.5°F from -4°F to 248°F.

Pros:

• The tightest published accuracy on this list, well under a fifth of a 5°F step
• Nearly instant, which helps if you check several pitchers in a row

Cons:

• Costs more. On ThermoWorks' own site in September 2026, the Thermapen ONE (opens in new tab) ran roughly two to three times the price of its ThermoPop 2 (opens in new tab), depending on whether it was on sale
• The speed matters less than it does in cooking. Wax in a pitcher cools over minutes, not seconds, so a 3-second reading and a 1-second reading tell you the same thing

Best for: makers pouring many pitchers per session, or anyone running careful side-by-side tests who wants the most margin. It's a real upgrade, not a necessity.

Infrared thermometer

Here's the good news for anyone who already owns one: it has a real job at the candle bench. It just isn't the job most people give it.

According to ThermoWorks' guide (opens in new tab), infrared thermometers read surface temperature only, "cannot see through glass or water," and are thrown off by shiny metal. On the emissivity table, polished aluminum is 0.05 and stainless steel 0.59. Glass is 0.92, close to the 0.95 (or 0.97) that fixed-emissivity guns usually use, according to ThermoWorks' infrared myth-busting article (opens in new tab). The laser, the guide adds, only helps you aim. It doesn't measure anything.

Pros:

• Instant, no-contact, nothing to clean
• Works well on high-emissivity surfaces like glass jars

Cons:

• Reads the skin of the wax, which cools before the wax underneath it
• Aimed at an aluminum or stainless pitcher, it reads the room's reflection, not the pot
• Can't tell you the fragrance-add temperature with any confidence

Best for: checking jars before a pour. CandleScience's IGI 4630 notes recommend preheating jars to help glass adhesion, and an infrared gun can tell you how warm a jar actually is in about a second. It's also handy for checking whether the surface you set candles on to cool is colder than the rest of the room.

Which candle thermometer to buy: the two-thermometer bench

Put the verdicts together and the setup is simple:

1. A digital probe goes in the wax. Choose one that publishes accuracy of ±1°F or better between about 100°F and 200°F. It's the only thermometer whose reading decides the fragrance add and the pour.
2. An infrared gun is optional, and it goes on the glass. Use it for preheated jars, the cooling surface, and the finished candle's top. Never point it at the pitcher.

If you're just starting out and can afford one tool, buy the probe. A glass clip thermometer can stay on the double boiler as a backup that doesn't need a battery, but once you have a probe, treat the probe's reading as the real one.

Dani ends up with exactly this. The probe settles her rough-top test in two batches, because a 5°F change is now clearly bigger than the thermometer's error. Her infrared gun, which used to point at the pitcher, now checks jars.

Check your thermometer before you trust it

A published accuracy spec describes a new thermometer on its best day. The one on your bench has been dropped, run hot and left beside the stove. FSIS gives a two-minute check (opens in new tab) that works for candle thermometers too:

1. Fill a large glass with ice cubes, add water and stir.
2. Put the probe at least 2 inches into the mixture, without touching the sides or bottom.
3. Wait at least 30 seconds. It should read 32°F.

FSIS's target is exactly 32°F. In practice, treat a reading within the thermometer's published accuracy as a pass: 31°F to 33°F for a ±1°F probe. If your thermometer publishes no accuracy at all, treat anything more than 1°F off as a fail for wax work, because your adjustments come in 5°F steps. A glass candy thermometer that starts at 100°F can't show 32°F, so it can't take this test. Check it with boiling water instead, or keep it as a melt-watcher and leave the 5°F experiments to a probe.

The ice test proves the thermometer at 32°F, which is a long way from 185°F. FSIS's boiling-water method (opens in new tab) gives a second check closer to working temperature. Steam burns, so keep your hand out of it or clip the probe to the pot:

1. Bring a pot of clean water to a boil. FSIS says distilled water is needed for true accuracy.
2. Put the probe at least 2 inches into the boiling water, without touching the sides or bottom.
3. Wait at least 30 seconds. At sea level it should read 212°F.

Treat this one as a rougher check. The ThermoPop 2's published ±1°F band stops at 208°F, and its spec table lists no accuracy between 208°F and 392°F, so there is no published tolerance at 212°F for that probe. (The Thermapen ONE's ±0.5°F band runs to 248°F and does cover it.) A reading within a couple of degrees of your local boiling point is reassuring. If it's off by more than about 2°F, treat that as a fail too. At high altitude, FSIS notes water most likely boils at least 2°F lower, perhaps as much as 5°F lower, and suggests asking your local Cooperative Extension office or health department for the exact boiling point where you live. If you haven't looked that number up, let the ice test decide pass or fail. A glass thermometer can't take the ice test, so look up your boiling point before trusting it, or keep it as a melt-watcher only.

If it fails, FSIS says to adjust it according to the manufacturer's instructions, and some models have a calibration nut or dial for that. If it can't be adjusted, FSIS says you may need to replace it. As a stopgap of your own until then, you can correct each reading by the error you measured: if it reads 34°F in ice water, subtract 2°F from every reading. Write that offset on the thermometer, and keep in mind that an error measured at 32°F may not be the same at 185°F. Check the offset against the boiling-water test too, and if the two errors differ, don't rely on the offset for 5°F pour changes.

Write the result down with the date. A probe that read 32°F in March and 34°F in September is drifting. Recalibrate or replace it before the drift shows up as a rough batch of candles.

Where the numbers go after the reading

A thermometer gives you one reading at a time. What turns readings into better candles is keeping them. The pour that finally smoothed Dani's tops only helps if she can find it again six weeks later, when she reorders wax and wonders why the new case behaves differently.

In Ardent Seller, a candle recipe keeps its steps next to its ingredients, so "heat to 185°F, add fragrance, stir two minutes, pour at 140°F" is part of the recipe instead of a sticky note. Each production run has its own notes field for what actually happened, like the pour temperature you hit, the room temperature, or a jar that went in cold. The thermometer can go in as an equipment item with a maintenance log, where each ice-water check is a dated entry. When a batch misbehaves, the temperature record sits next to the wax lot it came from.

If you want a paper version for the test phase, the free Candle Wick-Test & Fragrance Log has room for temperatures next to each burn test.

Pick the probe, check it in ice water, and give every number a place to live. Set up your first candle recipe in Ardent Seller free, with the fragrance-add and pour temperatures written into its steps.

Related reading

• Why Is My Candle Tunneling? — when a candle burns badly even though the temperatures were right, the wick is the next suspect, and this diagnostic goes through it.
• Candles per Pound of Wax — how many candles a pound of wax makes, so the batch you just learned to pour at the right temperature is also the right size.
• Best Scale for Soap Making — the same test applied to the scale: choose by the smallest amount you weigh, not the biggest batch.

Free resources

Two free downloads for the testing phase:

• Candle Wick-Test & Fragrance Log — record the fragrance-add and pour temperatures next to each burn test, so a 5°F change stays attached to the result it produced.
• Candle & Soap Fragrance Load Calculator — work out how much fragrance oil to weigh before the wax reaches 185°F, so the add goes in on time.

---

This article is provided for educational purposes only and does not constitute safety advice. Heating wax and fragrance oils carries fire and burn risks, and the temperatures above are supplier guidelines for specific products that may change. Follow your wax and fragrance suppliers' instructions and safety data sheets, and never leave melting wax unattended.