Every candle shop you visit will tell you to cure a soy candle for “at least four weeks.” Most shops will give you that number, shrug, and move on. It sounds like a ritual. It is not.
The two clocks you are not separating
What most people call “cure time” is actually three different processes running on three different clocks, all inside the same jar. Collapse them into one number and you either ship too early, or you hold inventory for no reason. Here is what is happening, and when.
Setting. Wax transitions from liquid to a solid crystal lattice. For soy, this happens inside 12 to 24 hours at 70°F ambient. Paraffin is faster. Coconut blends are slower, but the same order of magnitude. A candle that “looks set” at 24 hours is structurally set. Any time you add past that point is not setting.
Fragrance migration. Fragrance oil is not one compound. A typical FO blend is 60 to 220 individual aromatics, each with a different vapor pressure, chain length, and affinity for the wax matrix. When you pour at 135°F with 8% FO, the molecules are dissolved in a way they will not hold at room temperature. Over the first two weeks, the lighter aromatics (top notes: citrus, aldehydes, small esters) redistribute, and some of them escape through the jar headspace. That is why a fresh candle smells loud and a two-week-old candle smells rounder. Cutting cure at day 5 locks in the top-heavy profile and loses the base. This is the single biggest reason a “fast-cured” batch throws poorly on hot burn — the base loaders are the ones that carry the scent at burn temperature, and they are the last to stabilize.
Crystal annealing and container bonding. This is the piece nobody talks about. As soy cools from pour temperature down to room temperature, the crystal lattice shrinks. Soy has a higher coefficient of thermal expansion than paraffin, so it pulls away from the glass more aggressively. If your pour temperature was at the upper end of the window and your room was warm, the wax will contract by roughly 2 to 3% over the first 72 hours, and a small amount more over the next few weeks as the crystal lattice settles into a lower-energy structure. The wax-container bond is not formed at pour. It is formed as the wax slowly gives up that last fraction of contraction. A 7-day candle will pop out of the jar at 80°F in a truck. A 28-day candle will stay put. This is not a quality difference. It is a shipping and return-rate difference.
What actually happens in week 3
Most of the cracking and cold-splash defects we see from small producers show up in the day 10 to 21 window. The reason: the crystal lattice is doing its final reorganization. If the pour temperature was 5°F above the ideal window for your specific wax, or if the room temperature fluctuated by more than 4°F during the first 72 hours, the wax has built up internal stress that a 24-hour burn test will not reveal — the burn test is a 1-hour event, and the stress hasn’t had time to express itself. By day 10 to 14, the reorganization happens, and the stress releases as a visible crack, a cold splash, or a wavy top. This is why a batch that “passed” at day 3 and was shipped at day 7 can arrive cracked at the customer. The defect was already in the wax. It just hadn’t moved yet.
The practical consequence: your burn test schedule should not be one test at day 24. It should be a burn at day 7, a burn at day 14, and a burn at day 28, and the three should be compared. A candle that throws well at day 7 and poorly at day 28 has a fragrance migration problem — likely the load is at the ceiling for that wax, or the FO blend has too much volatile content. A candle that throws well at day 7 and cracks at day 21 has a pour-temperature or cooling-rate problem, not a fragrance problem. Different failure, different fix. You cannot diagnose either with a single day-1 test.
The numbers that matter
For a standard soy container candle at 8% fragrance load, 70°F ambient, and a clean pour in the 125°F to 135°F window:
- Minimum cure for a stable product: 14 days. Below that, you are shipping wax that is still contracting, with base notes still escaping.
- Target cure: 28 days. This is when the cold throw stabilizes, the hot throw stabilizes, and the container bond is complete.
- Coconut-soy blends (40/60 or 50/50): extend to 35 to 42 days. Coconut wax has a slower crystal-structure settling, and the fragrance load that works at day 28 in soy will often need an extra week in coconut.
- Paraffin: 7 to 14 days. Paraffin crystallizes to a stable structure much faster, and its fragrance migration is gentler. If you are curing paraffin for 4 weeks, you are holding inventory for nothing.
Cure is not a magic number. It is the time required for the slowest of the three processes to finish. Whichever one takes longest is your floor.
What to actually measure
If you are not already doing this, stop guessing and measure two things on three days. It takes 20 minutes per session and it will tell you more about your line than any “cure for four weeks” instruction ever will.
Cold throw. Unlit candle in a sealed 10-square-foot box for 30 minutes. Sniff at 30 seconds, 2 minutes, and 5 minutes. You are looking for the profile to be consistent — top, mid, base all present and balanced. A day-7 candle that is loud and bright is not the same candle as the day-28 version. Write down the difference.
Hot throw. Full melt pool, then off-flame. Smell at 10 seconds, 1 minute, and 5 minutes after the flame is out. This is where you detect base-note failure. If the day-28 candle is noticeably weaker than the day-7 candle on hot throw after 5 minutes off-flame, your base notes are escaping during cure and your load or your FO blend needs work, not more cure time.
The pattern you are looking for is: cold throw roughly flat from day 14 to day 28, hot throw slightly stronger at day 28 than at day 7, and no visible structural change (no new cracks, no new cold splash, no wax pop-out) from day 14 to day 28. If all three hold, your line is stable and 14 days is your true minimum. If any one fails, you have a specific problem and you know where to look.
The cure-time questions we actually get asked
“Can I speed up cure with heat?” No. Holding a candle at 90°F for 48 hours to “pre-age” it will accelerate fragrance escape, not stabilization. You will get a candle that throws worse on day 7 than one that cured at 70°F for 14 days. If you need faster turnaround, change your pour window, not your cure temperature.
“Do I need to cure unscented candles?” No. An unscented soy candle is structurally stable at 24 to 48 hours. The 28-day number is for fragrance migration, which does not exist in an unscented pour. If you are holding unscented inventory for 4 weeks, that is money you are not collecting.
“My customer says their candle is cracked. Did I not cure long enough?” Probably not. Cracking is a pour-temperature or cooling-rate defect that was present at the time of pour and expressed itself during transit. Cure time will not fix it, and it will not prevent it. Fix your pour window and your cool-down ramp, and the cracking stops. Adding two weeks to cure will not.
“Should I cure covered or uncovered?” Covered, after day 2. After the wax has set (24 hours), covering the jar slows the top-note escape and gives you a more consistent cold throw across the batch. This is a real, measurable difference — a batch cured uncovered will have a noticeably brighter top in the first 10 minutes of burning than a batch cured covered, and the covered batch will have better base notes at 30 minutes. If your product is meant to have a strong top-note opening, cure uncovered for the first 48 hours, then cover.
Cure is not a waiting game. It is three processes with three clocks, and the slowest one is your floor. Measure it, know where your line sits, and stop holding inventory you do not need to hold.