A new study being published in the November 2025 issue of the journal Industrial Crops and Products shows that hyperspectral imaging of cannabis fan leaves can accurately predict the cannabinoid content of mature flowers weeks before harvest.

Researchers from the University of Adelaide and the Australian Research Council Centre of Excellence in Plants for Space tested two cannabis cultivars under seven different lighting conditions. They used a handheld hyperspectral device to measure fan leaf reflectance early and late in flowering, then compared the data with harvested flower cannabinoid content. Machine learning models trained on the reflectance data achieved strong predictive accuracy, with R² values up to 0.89 for CBD, 0.77 for THC, and 0.8 for total cannabinoids.

Unlike traditional methods such as HPLC or GC-MS, which require destructive sampling and costly lab work, this technique allows for rapid, non-destructive, in situ analysis of intact leaves. The ability to forecast cannabinoid levels weeks before harvest offers significant benefits to growers and breeders. Industrial hemp farmers, for example, could identify plants at risk of exceeding legal THC thresholds and remove them before jeopardizing an entire crop. Medical cannabis operators could use the approach to prioritize high-yielding plants, shorten growth cycles, and reduce production costs.

The study also found that hyperspectral leaf data could distinguish between cultivars and lighting treatments, suggesting applications for germplasm classification and breeding programs. Predictions were reliable not only for major cannabinoids such as THC and CBD but also, in some cases, for minor cannabinoids like CBGA and CBCA.

Researchers emphasize that the method’s non-destructive nature sets it apart from earlier attempts at spectral prediction, which required sacrificing plant tissue. By using handheld equipment, growers and regulators could conduct on-site assessments without disrupting plant development.

The authors conclude that fan leaf hyperspectral reflectance could become a practical tool across both industrial and medicinal sectors, improving efficiency, quality, and compliance while reducing risks tied to cannabinoid variability.