The use of tetrahydrocannabinol vapes (THC vapes) has become widespread across diverse populations. For example, past month use prevalence was found to be 12.8% among 13–17-year-olds in New England, ¹ and 21.3% among electronic nicotine delivery system (ENDS) users aged 11-18 in a US nationally representative sample. ² This sustained level of THC vaping is particularly notable following the 2019 outbreak of vaping-associated lung injuries, where most affected individuals reported using THC-containing products. ³

For many years, laboratories have used standardized puffing regimens to assess toxicant emissions from tobacco products. ^{4– 6} Because toxicant yields are highly sensitive to puffing parameters, ^{7, 8} knowledge of device-specific puffing behaviors is essential for accurate emissions testing. However, such information for THC vaping devices remains limited. Emerging studies have either applied ENDS puffing regimens, ⁹ or used protocols developed for cannabis joints, ¹⁰ yet it remains unclear whether these regimens reflect actual use patterns for THC vapes.

To address this gap, we examined puff topography associated with THC vaping by analyzing publicly posted YouTube videos. We focused primarily on puff duration, while also documenting device types, liquid composition, and notable user behaviors.

We identified THC vape review videos and extracted information regarding device type, liquid composition, user behaviors, and puff duration.

In typical videos, reviewers unboxed and presented the device, took one or more puffs, and described the effects experienced. Devices were either pre-filled cartridges designed for external batteries or disposables combining battery and cartridge (see Figure 1). Reviewers prioritized describing cannabinoid and terpene composition, with delta-8 THC being most frequently reported. Other THC types mentioned included THCP, THCA, THCO, and delta-10.

Presenters often took short, rapid succession clearing puffs before inhalation. In disposable devices, several attempted the “blinker challenge,” drawing until the device’s LED light blinked to signal the puff duration limit. They frequently commented on vapor harshness and expressed a preference for smooth aerosols, describing flavors as “sweet,” “earthy,” or “cannabis-like.” Coughing after inhalation was common. After vaping, presenters assessed the intensity and onset speed of the psychoactive effects. Of the 98 individuals observed, 87 appeared to be male.

Across the 100 videos, 456 puffs were analyzed. Puff durations ranged from 0.12 to 18.01 seconds, with a mean of 3.68 seconds. The 25th percentile puff duration was 1.85 seconds, the median was 2.94 seconds, and the 75th percentile was 4.39 seconds. In the subset of 25 randomly selected videos, the second rater recorded a mean puff duration of 3.85 seconds compared to 3.68 seconds for the original rater; paired t-test analysis showed no statistically significant difference ( p = 0.7). Table 1 summarizes puff durations by device type. No statistically significant difference in puff duration was found between cartridge and disposable products ( p = 0.1). The raw data supporting these findings, including puff duration measurements from 100 publicly available YouTube videos and the reanalysis of 25 randomly selected videos by a second rater to assess measurement reliability, are available in the Mendeley Data repository. ¹¹

This study was conducted to address the lack of empirical data regarding puff duration during the use of cannabis vaping products, a critical factor for laboratory toxicant emissions testing. A secondary aim was to explore product attributes valued by users. By analyzing 100 YouTube videos of cannabis vape product reviews posted between 2018 and 2024, we were able to characterize puff topography and user preferences in a real-world context.

Our findings revealed a mean puff duration of 3.68 ± 2.66 seconds across 456 puffs. These results align closely with prior research. ¹² Self-reported puff duration data from an online survey completed by 557 respondents (75% identified as male), with a mean puff duration of 5.1 ± 2.5 seconds among the 293 individuals who vaped cannabis liquids. In a separate study, Smith et al. (2024) ¹³ directly measured puff topography from 9 participants over two weeks using a cannabis oil vaping device (Gram1™) equipped with an integrated puff counter and puff duration sensor, finding individual mean puff durations ranging from 2.20 to 6.95 seconds, with a population mean of 4.06 seconds. Importantly, the mean puff duration identified in our study is also comparable to those documented for users of electronic nicotine delivery systems (ENDS). For instance, Hua et al. (2013) ¹⁴ reported a mean puff duration of 4.3 ± 1.5 seconds in a sample of 64 individuals, and clinical studies of ENDS users have documented mean durations between 3.5 ± 1.4 seconds ¹⁵ and 4.16 ± 1.06 seconds. ¹⁶

These similarities suggest that existing ENDS puffing regimens may offer a reasonable initial model for laboratory emissions testing of THC vaping products. However, further research tailored specifically to cannabis vapes is warranted to confirm this applicability, given differences in device design, liquid composition, and pharmacology.

A notable limitation of this study is the underrepresentation of female users among the video presenters: 87 of the 98 individuals appeared male. This demographic imbalance contrasts with recent survey data showing comparable past-month THC vape use rates between women (39.9%) and men (38.1%), ¹⁷ suggesting limits to generalizability. Additional limitations include potential selection bias inherent to analyzing public video content, variability in video quality, and reliance on visual indicators of puff events.

Qualitative observations from the video analyses provided further insights into user priorities. Presenters frequently emphasized the importance of flavor characteristics, throat hit smoothness, and the rapid onset and intensity of pharmacological effects. Additionally, the “blinker challenge,” in which users attempt to trigger a device’s maximum puff duration limit, emerged as a common behavioral feature. This suggests that device features may function similarly to “Easter eggs”, hidden features that reward users with a sense of achievement for longer puffs and, consequently, higher drawn doses. This phenomenon warrants further investigation given its potential implications for user exposure and behavior.

In conclusion, this study contributes empirical data on THC vaping behaviors and emphasizes the need for laboratory protocols that accurately reflect real-world usage patterns.

This study did not involve human participants recruitment, or identifiable private information and therefore did not require ethical approval or informed consent. All data were obtained from publicly accessible YouTube videos that are freely available for viewing without restriction. No personal or identifiable information about individuals was collected, recorded, or analyzed. The study adhered to the ethical principles of using publicly available data for research and complies with F1000Research editorial policies.