How to Choose a Candle Wax

Updated 28 August 2026 · 15 min read

Wax is the fuel. Everything else in a candle — the wick, the fragrance, the vessel — is chosen around what the wax does when it melts, how much oil it will hold, and how it behaves as it sets.

This guide gives you the practical path first, then explains why the published figures for a single named wax vary from one seller to the next. Not similar waxes. The same product, with different numbers.

The short version

1. Pick a family by what you're making. Container, pillar, votive, melt — these want different things, and within paraffin especially, the grade matters more than the material. Typical published ranges put base and pillar grades at 3–4% fragrance, container and tart blends at 6–10%.[1]
2. Get the figures from the manufacturer, not the shop. Reseller listings for the same wax routinely disagree on melt point, maximum fragrance load and pour temperature. Find the manufacturer's technical data sheet or SDS.
3. Note your room. Pour temperature is calibrated to an ambient the supplier usually doesn't state. A workshop at 15 °C (59 °F) and one at 30 °C (86 °F) are not running the same process.
4. Set fragrance load below two ceilings. The wax's stated maximum, and your fragrance oil's stated maximum for candles. Use the lower.
5. Test in your own space, and cure before you judge. Vegetable waxes keep hardening for weeks after pouring. A candle assessed at two days is not the candle you'll sell.

Why the numbers disagree

On the wick guide, we found charts that conflicted because each was built on a different base wax and fragrance load. Wax is worse. Here the product is fixed — one manufacturer, one formulation, one name — and the published figures still diverge.

Ceda Serica: one wax, five spec sheets

Calwax make a coconut and apricot blend sold as Ceda Serica. Across five resellers:

Melt point appears as 125 °F (52 °C),[2] as 120–135 °F (49–57 °C),[3] and as 125–135 °F (52–57 °C).[4] Maximum fragrance load appears as 10%,[4] as 8–12%,[2] and as 8–10%.[3][5] Pour temperature is given as "try not to allow the wax to cool below 160 °F" (71 °C)[4] at one shop and as "mixing and pouring temperature 180–210 °F" (82–99 °C)[3] at another — a fifty-degree gap, with the upper end above the flash point of many fragrance oils.

Calwax's own safety data sheet gives neither figure. It states a congealing point of 123–135 °F (51–57 °C) by ASTM D938.[6] That's a different measurement, which we'll come back to.

Golden Brands 464: the load spread

464 is the most widely used soy container wax in the craft market. Melt point is reported consistently — 115–120 °F (46–49 °C) at Lone Star,[7] 113–119 °F (45–48 °C) at Scentivore.[8] Maximum fragrance load is not: 7–9%,[7] 8–10%,[9] and 10%.[8][10]

Pour temperature holds up better. Voyageur cite the manufacturer's recommendation as 125–145 °F (52–63 °C);[11] CandleScience and Scentivore both say 135 °F (57 °C),[8][10] which is the midpoint of that range. Convergent rather than contradictory — which shows the disagreements aren't inevitable.

IGI 4630: forty degrees on the pour

Melt point is agreed at 119–120 °F (48–49 °C).[12][13] Pour temperature is not. One shop gives a pouring range of 140–165 °F (60–74 °C);[13] another says pour at 180–190 °F (82–88 °C), then in the following line 160–190 °F (71–88 °C).[12] Lone Star's general guidance puts container single-pour blends at 150–160 °F (66–71 °C).[1]

And the one that settles it

Cargill make NatureWax C-3. On their own site, Cargill instruct you to heat the flakes to 160 °F (71 °C), and call it "the optimal pouring temperature to ensure beautiful, even candles."[14]

These gaps aren't random. Some are vocabulary. Some come from names that track processing rather than composition. Some are conditions the listing never states — including the room the guidance assumes and the room you're pouring in.

Your room is part of the recipe

The clearest evidence that pour temperature depends on ambient conditions comes from comparing hemispheres.

Pure Candle Supplies, an Australian wholesaler, give these instructions for their coconut soy blend: melt at 70 °C (158 °F), add fragrance at 50–52 °C (122–126 °F), pour at 45–48 °C (113–118 °F). They add that the wax "works best when poured at low temperature" and gives its best finish that way.[21]

CandleScience tell you to pour their coconut apricot at 170 °F (77 °C).[18] Windy Point say not to let Ceda Serica fall below 160 °F (71 °C).[4]

These aren't the same product — different blends from different makers — but the pattern holds: in these examples, the Australian pour sits roughly 45 °F below the American floor. Neither is an error. A candle setting in a 28 °C room loses heat far more slowly than one setting in a 20 °C room, and the whole point of pour temperature is to control how the wax passes through the range where it crystallises. ⚠️ The variable is thermal environment, not geography — an air-conditioned workshop in Brisbane and an unheated one in Ohio can swap places. The nationalities are a proxy for typical ambient conditions, and a poor one for any individual room.

Australian suppliers are explicit about this in a way American ones aren't. Wickii recommend their CocoSoy for "improved glass adhesion and smooth tops, particularly helpful in warmer Australian climates," and say they stock waxes "suited to different styles, climates and creative goals."[22]

The practical consequence: a published pour temperature assumes an ambient the seller usually doesn't state. CandleScience are a notable exception — their Lab Notes give the ambient room temperature their tests were run at.[23]

And one wax manufacturer publishes the relationship as a table. Cargill's handling sheet for NatureWax C-3 and C-6 keys pour temperature directly to room temperature — see the ambient table below. It is the strongest evidence on this page that the room is a process variable rather than a footnote, because it is the manufacturer saying so rather than us inferring it from disagreement.

You don't need to be in another hemisphere for this to bite. A Colorado workshop in January and a Florida workshop in August are further apart than Sydney and North Carolina.

The wax families

Four families cover almost everything, plus a fifth that most makers have never heard of.

Paraffin is refined from petroleum. It's the most forgiving wax on temperature, with a broad pour window and a glossy finish across most of the range.[25] But paraffin is not one material — grade determines almost everything about it, and a base grade and a container blend behave nothing alike.

Soy is typically based on hydrogenated soybean oil, often with additives or other waxes blended in — commercial soy waxes are formulated mixtures rather than a single material. Introduced for candles during the 1990s.[26] Michael Richards is often credited as its inventor, with Cargill's NatureWax C-1 as the original patent.[27] It's the dominant craft wax, and softer than the alternatives — soft enough that CandleScience warn 464 may be too soft to ship in summer heat.[10]

Coconut is rarely used alone — it's too soft to hold shape at room temperature, so it's blended, typically with soy for structure and sometimes with a small proportion of paraffin for throw.[4][18][19]

⚠️ "Coconut apricot" describes a marketing category, not a composition. Some blends contain paraffin. Others are explicitly paraffin-free — Wholesale Supplies Plus state of their coconut, apricot and soy blend that "there is no paraffin wax in this blend,"[53] and Candles and Supplies market theirs as "plant-based (NO PARAFFIN AT ALL)" with a 112–120 °F melt point.[54] The name does not tell you which. Only the ingredient declaration does — and as below, not always even then.

Beeswax is the temperature outlier. Relative to soft container soy and coconut blends, it's the one wax you pour hot: it has a narrow liquid window between melt and set, so jars need pre-warming to 50–60 °C (122–140 °F) or the wax skins on contact and gives a corrugated wet-spot pattern that's nearly impossible to fix.[25] Raw beeswax also holds propolis and pollen that scorch above 75 °C (167 °F).[25] CandleScience decline to give firm wick recommendations for it at all, saying that when, where and how the wax was obtained all affect how it performs.[28]

And a fifth: Fischer-Tropsch. Synthetic wax, made deliberately rather than recovered as a by-product. Chemically it's straight-chain saturated alkanes with high crystallinity and a notably narrow melting range, which is why it's sold on consistency and low shrink. A 2013 paper in the International Journal of Cosmetic Science notes the absence of aromatic and polycyclic aromatic compounds in Fischer-Tropsch wax[29] — the compounds most often raised as a concern with paraffin. It's uncommon in US hobby supply and more established in South Africa, the UK and Ireland.

What's actually happening in the jar

Soy and coconut waxes are hydrogenated triacylglycerols — the same class of material as chocolate and margarine. That matters, because it means their setting behaviour is governed by one of the best-studied problems in food science.

Three crystal forms, appearing in the wrong order

Triacylglycerols crystallise in three polymorphic forms: α, β′ and β. Stability, melting point and density all increase in that order, with β the most stable and highest-melting.[32]

But they don't form in that order. From a melt, the sequence runs α → β′ → β — the reverse of their stability. The least stable form appears first in a supercooled material.[32]

So in triacylglycerol systems crystallisation can proceed through less-stable forms before shifting toward more stable ones — α first, the softest and lowest-melting, converting over days and weeks toward β′ and β, which are harder and higher-melting. ⚠️ Commercial waxes are formulated mixtures, so the actual path depends on composition rather than following one fixed sequence. What you end up with is a mixture of forms, and which dominate is partly what a wax is formulated to control.

That is why vegetable wax keeps hardening after it looks finished. Armatage Candle Company describe it observationally — a soy candle poured on Monday is harder the following Sunday than it was on Tuesday, and most of the change happens in the first fourteen days but continues for weeks to months.[33] The polymorphic transition is the mechanism underneath.

This is why an under-cured burn test gives you the wrong wick. Test at two days and the wax is still largely in its soft α form. It melts readily, the pool looks deep, and you conclude you're over-wicked. A week later the same candle is harder and needs more heat. Size down on the early test and you'll have a tunnelling candle once the wax reaches its final state. See How to Choose a Candle Wick for what a correctly sized wick looks like.

Frosting and fat bloom share a mechanism

Frosting is best described as a crystallisation phenomenon related to the same polymorphism that produces fat bloom in other lipid systems — not as literally the same defect under another name. The analogy is useful because the mechanism carries over; it is not an identity.

β crystals are described as stable, compact and coarse, producing a granular texture.[34] α crystals are highly soluble and dissolve and recrystallise with temperature changes, growing larger crystals in the process.[32]

That accounts for what makers observe: frosting appears over time rather than immediately, and temperature swings make it worse.

It also explains the vocabulary. AAK call it "bloom/frost."[35] Cargill's resellers call it "fat bloom."[16] Makers call it frosting. All three describe the same phenomenon, and the term comes from food because the companies making vegetable wax are food companies using the same processes.

Chocolatiers solve this deliberately. Tempering means cycling temperature repeatedly to convert residual α before it can misbehave; without it, chocolate shows undesirable crystallisation.[36] Candle making rarely includes a tempering cycle; pour and cool replace it.

One consequence worth knowing: once the transformation has happened, the lower forms can only be recovered by completely remelting.[32] A frosted surface can be remelted with a heat gun. A frosted candle cannot be un-frosted.

The crystal network is why fragrance load has a ceiling

β′ crystals are the smallest — under 1 µm — and needle-like.[37] They interlock. Above a critical volume fraction they build a network extending through the whole volume, creating a solid with viscoelastic properties, and the ability of that network to entrap liquid oil depends on both the solid fat content and the crystal shape.[37]

So a wax's maximum fragrance load is partly governed by the carrying capacity of that solid network, along with the chemistry of the fragrance itself. Below the ceiling the oil is held; above it there is nowhere for the excess to go. ⚠️ It is not a single-cause outcome — addition temperature, the oil's own composition and polarity, and the wax's formulation all feed into where the ceiling falls, which is why two oils at the same percentage in the same wax can behave differently.

That's the mechanism behind what suppliers describe as symptoms. Lone Star note that the wax retains only so much fragrance oil and the excess separates instead of adding throw.[1] Fillmore Container call it syneresis or fragrance bleed — liquid accumulating on the top or bottom of the candle after cooling — and attribute it to excessive fragrance oil, or oil added at too low a temperature to bind.[38]

It also explains why β′ is the form formulators want. Small interlocking crystals give the network that holds oil. Coarse β crystals don't.

Cooling rate selects the form

Which polymorph you get depends on how fast the wax cools.[32] Fast cooling favours α, which means more subsequent transformation and more frosting.

That single fact unifies advice you'll find everywhere without explanation:

• Pre-warm your jars
• Don't cool candles on marble or granite, which draw heat out too quickly[10]
• Cover or box candles for two to three days[13]
• Bring containers to at least room temperature before pouring[38]

Every one of those slows the cooling rate through the crystallisation range.

It also resolves the apparent disagreement from earlier. Pouring cool into a warm room and pouring hot into a cooler one are both attempts to control the same cooling profile from different starting conditions — which is exactly what Cargill's ambient table prescribes, and why the split tracks thermal environment rather than hemisphere.

Fragrance load and temperature

Two questions, both with more disagreement than you'd expect: how much fragrance the wax will hold, and how hot it should be when you add it.

The load ceiling is the lower of two numbers

Every wax has a stated maximum. Every fragrance oil has a stated maximum for candles. Use whichever is lower — Lone Star put it plainly on their product pages: the fragrance load is the maximum the wax holds, and your oil's own IFRA ceiling may be lower.[7]

One misconception worth clearing up: IFRA's categories aren't a fragrance-load limit for candles. The twelve categories are ordered by skin contact, and candles sit in Category 12. IFRA's own guidance defines Category 12 as products not intended for direct skin contact, with minimal or insignificant transfer, and states that because skin exposure is negligible, "the concentration of fragrance ingredient is not restricted in the finished product."[40]

But restrictions are set per material, and some still bind in Category 12 through systemic toxicity rather than skin sensitisation. The number that governs your oil is on its Certificate of Conformity.[40] Check that, check the wax maximum, and use the lower.

You'll also find published maximums that exceed the manufacturer's own. Cargill recommend 6–9% for NatureWax C-3;[14] two resellers publish 10%.[15][16] Lone Star's guidance is that you can load up to the maximum but never past it[1] — which requires knowing whose maximum you're reading.

Exceeding it has a specific failure mode

The excess doesn't disperse. It separates — beading into droplets on the surface or pooling at the bottom.[38] Where excess oil collects as free liquid, fire-safety risk rises.

Downstream symptoms include sooting, an oversized or erratic flame, wax pulling down around the wick, wet spots, poor glass adhesion and shortened burn time.[38]

Fillmore Container add a nuance worth keeping: bleed can also come from fragrance added at too low a temperature to bind, not only from too much of it.[38] Same symptom, different cause.

And one belief worth retiring: more fragrance does not mean more throw. Past the wax's capacity it inverts.[41]

When to add it: everyone heats to the same temperature, then parts company

The interesting thing is what they agree on. Nearly everyone heats to 185 °F. The disagreement is whether fragrance goes in at the peak or after cooling — and that's a process difference, not a rounding gap.

CandleScience describe 185 °F as a "Goldilocks Zone": hot enough for the fragrance to fully bind with the wax, cool enough that no fragrance is lost.[10] Fillmore give the same temperature specifically to prevent bleed. The Flaming Candle's product-page reasoning runs the other way — add late, to preserve the scent.

So one camp optimises against a binding failure and the other against volatilisation. Neither is careless; they're protecting against different things.

And one supplier disagrees with itself. The Flaming Candle's general guide says 165–185 °F; their NatureWax C-3 product page says 160–165 °F. Those ranges share exactly one point. The product page is the more specific document and the one to follow for that wax — but you'd have to notice the discrepancy to know there was a question. "The Flaming Candle says X" isn't a citable position; the page is.

The Australian figure sits 60 °F below the American consensus, and the ambient explanation from earlier covers part of that — but not the peak-versus-cooled split, which is a genuine disagreement about what goes wrong first.

One ceiling worth knowing: The Flaming Candle caution that heating wax above 200 °F (93 °C) may cause discolouration.[15] Cargill's own handling sheet gives the same limit for NatureWax C-3 and C-6 — melt at 160–200 °F (71–93 °C), and "do not heat the wax above 200 °F," because wax held above it for long periods will discolour.[52]

The manufacturer publishes a room-temperature table

Everything above infers from supplier disagreement that the room is part of the recipe. One manufacturer states it outright. Cargill's handling sheet for NatureWax C-3 and C-6 says pour temperature "should be matched to ambient room temperature recommendations in the table below," and then gives the table:[52][14]

Note which way it runs. A warmer room takes a cooler pour, not a hotter one — because the room governs how fast the candle sheds heat, and a slow-cooling candle poured hot is the one that cracks and discolours. Across a 30 °F swing in room temperature the recommended pour moves by 25–35 °F.

This is the single best answer to "why do two suppliers give different pour temperatures for the same wax." Some of that spread is not disagreement at all — it is the same instruction, written for different rooms. It also means a pour temperature quoted with no ambient condition attached is an incomplete instruction, however precise the number looks.

⚠️ The sheet is not internally consistent, which is worth seeing rather than smoothing over. Step Four says to cool the wax to 120–165 °F before pouring; Step Five says to pour at 120–180 °F; the table's top row says 170–180 °F. A pour at 175 °F satisfies the table and Step Five and violates Step Four. This is the page's whole argument appearing inside a single manufacturer document — and it is a reason to treat the table as the specific instruction and the step ranges as the envelope.

Labels and claims

This section is about reading claims, not legal advice for labelling your own products.

Bourbon has a legal definition. Under federal labelling rules for distilled spirits, Title 27 of the Code of Federal Regulations specifies 51% corn, new charred oak containers, distillation below 160 proof.

Candle wax has no equivalent.

There is no US standard of identity for wax

The regulations that touch candles concern fire safety and warning labels — ASTM F2417 for fire safety performance,[42] ASTM F2058 for labelling, and the CPSC ban at 16 CFR 1500.17(a)(13), effective 15 October 2003, on metal-cored candlewicks containing more than 0.06% lead by weight in the metal, and candles containing them.[43] Note the shape of that rule: it's a lead threshold, not a prohibition on metal cores — zinc-core wicks remain legal.

There is no federal US standard of identity defining what a wax sold as soy, coconut or natural must contain.

To be clear about what that does and doesn't mean: general law still applies. The FTC prohibits deceptive advertising, safety data sheets are required, and IFRA standards govern fragrance materials for members. What's absent is a composition standard — a rule saying that "soy wax" must be a stated percentage of soy.

Which explains everything documented earlier in this guide. Golden Brands 464 is sold as all-natural soy while containing proprietary soy-based additives that let it pour hotter and hold more fragrance.[10] Ceda Serica is described by resellers as "made from renewable sources"[3] while its own safety data sheet refers to a hydrocarbon component.[6] Three waxes share the name coconut apricot with three different compositions.

One arithmetic point that follows, because it changes what a number means. Fragrance load and fragrance content are not the same figure. Ten grams of fragrance in 100 grams of wax is a 10% load — but it is 10 ÷ 110 = 9.09% of the finished candle's mass. Suppliers quote load; IFRA limits are expressed as a fraction of the finished product. The gap is small at 6% and grows with the load, and it is the kind of difference that decides whether a formula sits inside a limit or outside it.

The practical response: if you want to know what's in a wax, read the manufacturer's documents — but know what each one is for. The technical data sheet says how the manufacturer expects the wax to perform. The SDS is a hazard-communication document, written to tell you how to handle the material safely and what to do if it goes wrong. Neither gives the full formulation, and the SDS in particular is not an ingredient list — it discloses what regulation requires it to disclose, which for an unhazardous wax can be almost nothing. The product page is written to sell it.

Free-from claims, and what's actually in question

Wax listings carry a lot of absences. Gluten-free, paraben-free, sulfate-free, toxin-free, phthalate-free, non-GMO, vegan, cruelty-free, biodegradable, non-toxic.

Some of those describe things that were never candidates. Gluten-free wax is true of every wax ever made — wax contains no wheat, and the claim is borrowed from food marketing. Paraben-free and sulfate-free are similar: parabens are cosmetic preservatives, sulfates are detergents. Neither belongs in wax to begin with.

That's worth naming plainly, and it generalises: a "free-from" claim is informative only when the excluded substance was a plausible ingredient. Otherwise it tells you nothing about what is in the product. Paraffin-free on a wax is a real claim, because paraffin is a plausible ingredient and often present. Gluten-free is not.

Non-toxic and toxin-free have no definition at all, and no certifying body. They aren't verifiable claims in either direction.

Two are meaningful:

Vegan matters, because beeswax isn't. It's why AAK describe their waxes as "completely free of animal-derived substances" rather than simply asserting the word.[35]

Non-GMO is the one that carries real information and gets the least attention. USDA reported 96% of US soybean acreage planted with genetically engineered herbicide-tolerant seed in 2024 and again in 2025.[44] Non-GMO runs about 4% of total acreage, and the food-grade portion — roughly 2% in 2025 — is largely contracted, commands a premium of around $2.53 per bushel, and goes mainly to tofu, soy milk and miso, with Japan the largest importer.[45]

So a soy wax should not be assumed non-GMO unless the manufacturer makes and substantiates the claim — certification is one route to that, but not the only one, and the base rates above are why the assumption runs the way it does. AAK's brochure lists "Sustainable and Non GMO options available"[35] — which tells you the standard product is neither.

Whether that matters to you is a separate question. But it's a fact with a primary source, and almost nobody states it.

Testing

Everything above is a starting point. This is where you find out what's true in your workshop.

Cure before you judge anything. This is the single most common way to test correctly and still get the wrong answer. Vegetable wax keeps hardening for weeks, so a candle assessed at two days behaves differently from the one you'll sell. Cure your test candles for the same length of time you'll cure production — and judge your wick on those, not on a fresh pour.

Published cure times vary widely, and some of that variation is real product difference: CandleScience claim one to two days for their coconut apricot,[18] two weeks for 464.[10] Ceda Serica resellers specify fourteen days.[5] Australian coconut soy guidance says cool 24 hours, cure 48 minimum[21] — though a label minimum isn't necessarily a full polymorphic settle.

Record your conditions. Room temperature, container temperature, pour temperature, fragrance load, cure time. Without them you can't tell which change produced which result — and you can't compare your results to anyone's published figures, because they're testing in a different room.

Change one thing at a time.[41] Wax, wick, fragrance, load and pour temperature all interact. Change two and you've learned nothing.

Watch for the failure modes with known causes. Wet spots and pull-away point to shrinkage and cooling rate. Frosting points to crystal transformation and temperature swings. Oil beading or pooling points to load exceeding the wax's capacity, or fragrance added too cool to bind. Sinkholes and craters are often a pour temperature problem — CandleScience suggest adjusting in 5 °F (3 °C) increments to find what works in your space.[50]

And expect the wax itself to vary. Moisture content differs between cases of the same wax, and you may see cloudiness before pouring.[38] Burn rate varies enough between production lots that AAK sell consistency as a product feature — one of their case studies describes a customer who no longer needed to change wicks based on the burn rate of each incoming lot.[35] If a batch behaves differently and nothing in your process changed, the wax is a candidate.

⚠️ And comparing two waxes by name compares nothing. Changing the wax changes the wick you need, so a straight swap tests the pair, not the wax — if wax B burns worse in a candle wicked for wax A, you have learned about that combination and nothing about wax B. Which wax is better and which wax-wick combination is better are different experiments, and only the second one is usually run. To compare waxes, re-optimise the wick for each and compare the optimised candles; see the wick guide's testing section for how to run that without changing two things at once.

What to take away

There is no single right wax, and the published numbers are less precise than they look. The same product carries different figures at different shops. Pour temperatures assume a room nobody names. Names track processing rather than composition, and no standard defines what any of them must contain.

That's not a reason to distrust the numbers. It's a reason to know whose they are, what conditions produced them, and where your own workshop differs.

Get the manufacturer's documents. Note your room. Cure properly. Change one variable at a time.

And when a wax behaves in a way the chart didn't predict — that's the normal case, not the exception.

About · Contact · Privacy Policy · Terms of Use · Editorial Standards · For brands