Study: three different ways to smoke a joint – all pull more smoke than the cigarette standards

For the first time, researchers have measured how cannabis users actually pull on a joint in real-world settings. At social gatherings in five US states, 62 people smoked a one-gram joint with a flowmeter attached. The result was three distinct puff profiles – and all of them sit above the cigarette standards that laboratories have used for decades.

62 people were measured at social gatherings in five states

The measurements were taken between September 2024 and June 2025 in California, Massachusetts, New York, Nevada and South Carolina. Participants smoked anonymously and were asked to smoke exactly as they normally do, using one-gram joints made from various flower materials, including hemp and type II cannabis.

Each session lasted at least three minutes, and the CannaFlo flowmeter logged ten data points per second. Of the 62 recordings, eleven were excluded because of insufficient puff data, pressure-drop problems or baseline irregularities, leaving 51 people in the analysis.

Three profiles: intense, slow and double

The most common profile is called intense and accounted for 57 percent of the participants, 29 people. It is characterised by a fast, forceful draw that reaches its highest flow rate after just 0.8 seconds and then fades. The averages were 2.3 seconds per puff, a peak of 41.7 millilitres per second and 59.3 millilitres of smoke per puff.

The slow profile was the least common at 20 percent, ten people. There the puff lasts 4.7 seconds on average and peaks lower, at 22.4 millilitres per second, but holds a steadier level instead. That paradoxically delivers more smoke in total: 72.5 millilitres per puff.

The third profile, the double puff, accounted for 23 percent, twelve people, and consists of two rapid puffs a couple of seconds apart. Counted together, the puff lasts 3.3 seconds and delivers 76.8 millilitres of smoke per puff, the largest volume of the three.

"It is really a difference between a fast, intense pull, almost like sipping boba through a straw, and a long, slow pull that is more like sipping a thick milkshake", says co-author Riley Kirk of the Network of Applied Pharmacognosy in an interview with High Times.

The cigarette standards draw less smoke

When laboratories analyse cannabis smoke they use machines programmed with puff parameters borrowed from tobacco research. The Health Canada cigarette standard corresponds to 55 millilitres of smoke per puff over two seconds, peaking at around 27.5 millilitres per second, and the ISO standards sit in the same range.

All three cannabis profiles are at or above 55 millilitres. The intense profile is the one that most resembles the cigarette pattern, while the double puff reaches almost 40 percent more smoke. According to the authors, that means existing laboratory research on cannabis smoke rests partly on behaviours that do not match how cannabis is actually smoked.

The chemistry of the smoke has not been measured yet

The study maps behaviour, not content. It says nothing about how much THC, terpenes or breakdown products actually reach the user. The authors hypothesise that puff length and flow may affect the temperature of the ember and thereby the chemistry of the smoke, but that remains to be measured.

Limitations and funding

The report is a white paper without peer review, and the authors are open about its limitations. The 62 participants were recruited at social gatherings rather than through random sampling, and paper, filters, grind, moisture and the age of the joint varied between sessions. The measurements also cover at least three minutes per person, not the whole joint, and the measurement device itself may have influenced how participants smoked.

The Science of Smokability research programme is funded by industry sponsors, including Abstrax, Boveda, Cannatrol, Curaleaf, Custom Cones, Day Savers, Ganjier Group, Hemp Mine and Plant Life Co. Two of the co-authors work for Cambustion, the company that makes the flowmeter.

Source

From User Behavior to Laboratory Standards: Developing Cannabis Smoking Machine Protocols from Real-World Puff Topography Data