Importance: While a diverse array of cannabis products that may appeal to youth is currently available, it is unknown whether the risk of persistent cannabis use and progression to higher frequency of use after experimentation differs among cannabis products.

Objective: To estimate the comparative relative risk of experimental use of 5 cannabis products on use status and frequency of use among adolescents during 12 months of follow-up.

Design, Setting, and Participants: In this cohort study, data were collected from 3065 adolescents at 10 high schools in southern California, with baseline data collected in spring 2016, when students were in 11th grade, and 6-month and 12-month follow-up surveys collected in fall 2016 and spring 2017, when students were in 12th grade. Analyses, conducted from April to June 2019, were restricted to 2685 participants who were light or nonusers of any cannabis product (ie, ≤2 days in the past 30 days) at baseline.

Exposures: Number of days of use of each cannabis product (ie, combustible, blunts, vaporized, edible, or concentrated) in the past 30 days at baseline (ie, 1-2 vs 0 days).

Main Outcomes and Measures: Past 6-month use (ie, yes vs no) and number of days of use in the past 30 days at 6-month and 12-month follow-ups for each product.

Results: Of 2685 individuals in the analytic sample, 1477 (55.0%) were young women, the mean (SD) age was 17.1 (0.4) years, and a plurality (1231 [46.6%]) were Hispanic individuals. Among them, 158 (5.9%) reported combustible cannabis use on 1 to 2 days of the past 30 days at baseline, 90 (3.4%) reported blunt use, 78 (2.9%) reported edible cannabis use, 17 (0.6%) reported vaping cannabis, and 15 (0.6%) reported using cannabis concentrates. In regression models adjusting for demographic characteristics and poly-cannabis product use, statistically stronger associations of baseline use with subsequent past 6-month use at the 6-month and 12-month follow-ups were observed for combustible cannabis use (odds ratio, 6.01; 95% CI, 3.66-9.85) and cannabis concentrate use (odds ratio, 5.87; 95% CI, 1.18-23.80) compared with use of blunts (OR, 2.77; 95% CI, 1.45-5.29) or edible cannabis (OR, 3.32; 95% CI, 1.86-5.95) (P for comparison < .05); vaporized cannabis use (OR, 5.34; 95% CI, 1.51-11.20) was not significantly different from the other products. In similarly adjusted models, we found the association of cannabis use at baseline with mean days of use at the 6-month and 12-month follow-ups was significantly stronger for cannabis concentrate than for other cannabis products; participants who had used cannabis concentrate on 1 to 2 of the past 30 days at baseline (vs 0 days) used cannabis concentrate a mean of 9.42 (95% CI, 2.02-35.50) more days in the past 30 days at the 6-month and 12-month follow-ups (P for comparison < .05).

Conclusions and Relevance: Cannabis control efforts should consider targeting specific cannabis products, including combustible cannabis and cannabis concentrate, for maximum public health consequences.