Cone or straight joint? You’ve got a side. Everybody’s got a side. Nobody’s ever had proof. So a research team packed a dozen of them with the exact same weed, smoked every one on a machine programmed to puff more like a cannabis consumer than a cigarette smoker, and weighed what came out the other end. Here’s the kicker: you were both right.
You know how this goes. One guy loves a pre-rolled cone. The next guy swears by a straight cylindrical joint and won’t hear otherwise. And it never gets settled, because the only thing anybody’s ever brought to the fight is how sure they sound. It’s a two-in-the-morning debate that’s dead by lunch.
Not anymore. There’s finally data. A research team led by Drs. Riley Kirk, Miyabe Shields and Allison Justice, with Twinkle Paryani, took one jar of flower and split it between seven pre-rolled cones and five cylinder joints. Then they burned every single one on a lab machine set to a cannabis-specific puffing profile instead of the old cigarette settings, caught the smoke on a filter, weighed it, counted the pulls, and clocked how hot each one ran. And the big result is almost too fair: the machine detected no significant difference in the total THC or terpenes captured from either format. The difference was in how that material got divided across each puff.
Similar totals. What changes is how they arrive. The cone loads more into every puff and finishes in fewer pulls. The straight joint takes that same weed and stretches it out, thinner, over more pulls. Neither one gave up a bigger total. They’re just two different speeds for smoking the same gram, and now somebody’s actually got the numbers.
The Cone Packs More Into Each Puff. Here’s the Why.
Quick translation: “more per puff” means more THC caught by the machine on each draw. Nobody wired up a person to measure the actual high. What they measured was the smoke coming off the joint, and on that score the cone came out ahead per pull.
The likely reason is the shape. Look at the burning end of each. A pre-rolled kingsize cone is fat at the tip, a touch over 10.5 millimeters across. A cylinder joint is skinnier, just under 8. Doesn’t sound like much until you turn it into burning surface: the cone’s lit end has roughly 78% more area. The authors point to that geometry as the main driver, more surface, potentially more material combusting on each draw.
The numbers: the cone delivered 2.40 milligrams of THC per machine puff against 2.02 for the cylinder, plus more of the big flavor-and-aroma terpenes per puff too. Both gaps cleared statistical significance. And because the cone works through its weed faster, it’s done in fewer pulls, about 20 on average versus 23 for the cylinder. Same gram, roughly 12% fewer pulls to finish it. Fewer, denser puffs versus more, lighter ones. That’s the trade.
The Real Surprise: Cones Were Way More Consistent
Here’s the finding that might actually sway how you pick. The cone results clustered tightly. The cylinders spread much wider by comparison.
On total smoke resin, the cones landed at a coefficient of variation of 2.9%, a measure of how tightly the samples clustered, against 16.5% for the cylinders. In plain terms, cone after cone came out close to the same, while the cylinders scattered a lot wider. Several of the cannabinoid and terpene measures followed the same general pattern, notably tighter among cones.
Here’s what the study can’t tell you, though: why. The cones were packed and tamped by the researchers; the cylinders were formed on a RAW rolling machine with the flower loaded by hand. So the difference could come from the geometry, the packing method, the machine rolling, or some mix of all three. The study wasn’t built to separate them. What it shows is that in this small batch, the cones came out more uniform. Why is the next question.
Finally, a Weed Study That Doesn’t Smoke Like a Cigarette
Here’s the part that sets the study apart from a lot of what came before it. For decades, the machines testing cannabis smoke ran on cigarette settings, small puffs on a rigid timer, built around how a lab decided a tobacco smoker puffed fifty years ago. Nobody smokes a joint like a cigarette.
This team programmed the machine with a cannabis-specific puff profile they say was built from real-world consumer data, bigger draws, a different rhythm, instead of leaning on the old cigarette protocol. Then they ran everything through a lab-grade smoking machine and caught the smoke on a filter for analysis. It’s a small study, twelve joints, one strain, and the researchers say plainly that the trends need a bigger batch before anyone treats them as settled. But the puffing setup is a real step toward measuring cannabis smoke on cannabis terms rather than tobacco’s.
The Science, for the Nerds
If you don’t care about p-values, you already got the answer up top. For everyone else, here’s the full teardown.
The team burned seven RAW Classic Kingsize Cones and five RAW Classic cylinder joints, each packed with one gram of the same ground flower, the strain Peaches and Cream, testing at a total potential THC content of 14.85%. Everything ran on a Cambustion SCS smoking machine under a cannabis-specific puff profile (a 41.7 mL/s peak flow rate and a 59.31 mL puff volume), with particulate captured on a Cambridge filter pad per the ISO 3308:2000 standard. Significance was tested with Welch’s two-sample t-test at p < 0.05. This was an exploratory study: seven cones, five cylinders and only one cultivar. The report also ran numerous comparisons without describing a correction for multiple testing, so the significant findings need replication.
On total yield, the two formats tied statistically. The machine captured 48.96 mg of THC per cone versus 46.83 mg per cylinder (p = 0.43), 1.67 mg versus 1.46 mg of major terpenes (p = 0.18), and near-identical minor terpenes (p = 0.93). Transfer efficiency from flower to filter, how much of the payload made it into the collected smoke, came in at 32.97% for cones and 31.54% for cylinders, both in line with prior published work. And a nonsignificant difference at this sample size means the totals looked similar, not that they’re proven identical; the study didn’t run an equivalence test. The one cannabinoid that separated the formats was CBN, modestly higher in cones (0.46 mg versus 0.40 mg, p = 0.028). The authors suggested combustion differences may explain it, though the temperature picture was mixed: cones ran hotter through the middle of the smoke and held a higher temperature before the first ash break (138 °F versus 121 °F, p = 0.002), while cylinders trended toward higher peak temperatures at the very end (p = 0.068, not significant).
The per-puff numbers are where significance showed up: THC per puff (p = 0.0015) and major terpenes per puff (p = 0.009) both cleared the bar, as did puff count to completion. The team also ran the pooled ash through elemental analysis (ICP-MS) and flagged differences in aluminum, sodium and selenium between formats, but they had to combine samples to get enough material to test, so those are a thread to pull later, not a finding. As the researchers put it, that data needs a bigger scale and repeat runs before anyone leans on it.
So, Cone or Straight?
Nobody wins, and that’s the honest answer. Under these machine conditions, the formats produced statistically similar total captured THC and terpenes from the same gram. The cone concentrated more into each puff and finished in fewer pulls. The cylinder spread similar totals across more.
The study can’t tell you which one feels better, burns better in your hand, or makes for a better night. Nobody measured that, and honestly that’s the part you’re deciding for yourself anyway. Want fewer pulls with more THC captured in each puff? The numbers lean cone. Prefer more pulls from the same gram? Go straight.
Same weed, similar totals, different pacing. The oldest argument at the session finally has somewhere real to start, and it turns out the format changes the pace of the ride more than the size of it. Which is just a long way of saying what we told you up top. You were both right.
Editor’s note: This article is based on a May 2026 data report, “HBI Innovations Smoke Study: Cones vs. Cylinders,” conducted by researchers Riley Kirk, Miyabe Shields, Allison Justice and Twinkle Paryani of Kirk and Shields LLC, with laboratory support from Gobi Labs and the Clemson University Agricultural Service Laboratory. The study tested RAW-brand cones and rolling papers. The report is preliminary and has not been peer reviewed.
