You have seen the advice. It is everywhere, it comes from serious shops, and it is slightly wrong. "Burn for two hours." "Burn for three to four." "Never burn a candle for more than four hours." Read a dozen candle sites and you get a dozen different clocks, all of them telling you to watch the time instead of the wax. That is the real mistake hiding inside the whole category of burn-time advice: nobody tells you what the clock is actually measuring.

I have poured and tested enough candles to know the answer is boring and specific. The flame does not read your watch. The wax is a hydrocarbon, and hydrocarbons do not care about your evening plans. What happens to your candle is determined by temperature, distance, and a reaction that starts the moment you strike a match. So let us look at what a burn session actually does to the three things in your jar — the wick, the wax, and the glass — and where each one breaks.

The problem: a rule with no mechanism.

"Two hours," "three hours," "four hours." These numbers sound like they come from a study. They do not. They come from small jars with small wicks burning low-viscosity waxes in a specific room, and then every other brand copied the number without the conditions behind it. A 2.5-inch jar with one cotton wick in a coconut blend may genuinely need two hours to melt across its full width. A wide 5-inch bowl wicked for long melts may take four to reach the edge. The same "rule" is right for one candle and actively harmful to another — too short for the big one (permanent tunnel), too long for the small one (overheated glass, dead wax).

Here is the mechanism the rules never explain. A candle burns in the gas phase. The wick carries liquid wax up by capillary action; the vaporizes it; the gas burns; the heat of that flame melts more solid wax into the pool. The pool is not a puddle, it is the machine. Everything you care about — scent, clean burn, glass survival — runs off the temperature of that liquid surface, which is set by the flame and the distance the heat has to travel. When you set a timer, you are substituting a guess for a measurement you could have taken with your eyes.

What long burns actually do, in order of when they happen:

First: the wick mushrooms. The burning tip is where unreacted carbon accumulates because the flame runs a little fuel-rich at its edge. Minute by minute a small carbon cap builds up — that "mushroom." It makes the flame taller and wider, which means hotter, which means more carbon builds. It is a feedback loop. A mushroomed wick is what you see in every soot complaint: black dust on the rim, flickering that looks like the candle is struggling, and a flame that is now too far from the liquid to pull wax up at the right rate. The flame starves at the bottom and burns dirty at the top. Trimming to a quarter inch before every burn is not a tip, it is the only reliable reset you have.

Second: the wax itself burns off its own fragrance. This is the one nobody prints on the label. Your fragrance oils are smaller, more volatile molecules than the wax around them. As the pool sits at burning temperature for a long session, those top-note molecules evaporate into the air — the part of your scent profile that made the candle interesting leaves the jar forever, and what remains is a duller, wavier base. This is why a candle that smelled sharp and bright in month one smells "muddy" in month two, even with the same burn habits. It is not the storage that did it, it is the hours you let it run. The pool is a slow distillation plant, and you are the customer for the cheap fraction.

Third: the glass gets hot, and glass is the weakest link in the whole system. A burning candle pool runs warm enough that the glass above it climbs well past comfortable-to-touch in a long session. Glass tolerates heat, but it tolerates heat slowly and evenly. What kills it is the gradient — a hot zone near the wick and a cooler rim, held for a long time. That differential stress is what cracks a jar, usually along a scratch or a seam, sometimes with a hairline fracture you do not notice until it fails. The advice "stop burning when there is half an inch of wax left" is not a politeness suggestion. In the last inch of a jar, the wick sits close to the bottom glass, the heat has nowhere to escape but into the vessel, and the thermal stress on the glass peaks exactly when the candle is closest to the point of failure. A cracked jar with hot wax in it is the one bad night in the whole hobby.

So what do you actually do? The working rules, in priority order:

  1. Watch the pool, not the clock. The session is done when the melt pool reaches the glass on all sides — the wax has melted a full width, edge to edge. That is the signal. It may be one and a half hours in a small jar or four hours in a wide bowl. Your candle's real burn time is a property of its diameter, wick, and wax; the number on the label is someone else's candle.
  2. Never let it run out of wax. Keep at least half an inch in the jar. If you are down to the nub, stop. The glass is now absorbing more heat than the wax, and you are one more hour from a crack.
  3. Trim before every light. Quarter inch. The mushroom cap is the reason for the soot, the wandering flame, and the "my candle smells like a campfire" complaints. The fix is thirty seconds with a trimmer.
  4. One full melt, then stop. The pool that reaches the edge sets the "memory" for the next burn. Short, repeated sessions that never reach the glass are what build the tunnel — the deep hole in the middle with a wall of solid wax around it that the flame can no longer reach. A tunnel is a candle that has been slowly starved by its owner. If you have one, the fix is mechanical: re-melt the surface with a heat gun or a thick wick in the hole until the pool is level again, then burn to a full width.

The number, if you insist on one: for a standard 8-ounce jar in a room with normal airflow, two to three hours is the range where a properly wicked candle reaches its full pool without pushing the glass. Four hours is the outer limit, and only for jars that were designed for it — the diameter, the wick count, the wax blend all have to agree. Past that, you are no longer burning a candle, you are testing a jar, and the jar is the one losing.

The two-hour rule, the four-hour rule, the "burn until the melt pool is full" rule — they are not competitors. They are the same observation described at different precisions. The first is a safe default. The second is the ceiling. The last one is the actual thing, and it is the only one that holds across every size and every wax. Everything else is just the number somebody put on the label because the customer wanted a number.

If your candle is tunneling, sooting, or pulling away from the glass, the problem is almost never the wax. It is the session length, the trim, and the last half-inch of wax you were not willing to leave behind.