In clinical research, cannabis is often set up to fail. Federal restrictions privilege studying isolated molecules from cannabis over the plant itself. And even when cannabis research is done, scientists are often forced to use low-quality weed from the National Institute of Drug Abuse. Because of this, most of the useful medical data on cannabis comes from observational studies, where people consuming cannabis are asked about their experience. Researchers in Albuquerque recently analyzed an app-based survey of over 3,000 cannabis users. They found that higher THC levels predicted better symptom relief for most problems (including anxiety and depression, but not all kinds of pain). CBD levels were not related to symptom relief, on average. This format for research, however, is likely biased. THC provides an immediate and obvious effect, whereas CBD may confer its effects with more subtlety. We don’t know how people found the app, or even that the app developers gave all the data to the researchers, so these results might not be representative of the medical marijuana community in general. The scientists are forthcoming about limitations of the study. However, they only report the correlation between cannabinoids and symptom relief for three conditions, not all 27 conditions surveyed. And because the data are owned by the app developer, we cannot interpret the results beyond what the authors have chosen to present. Despite these caveats, the study highlights a significant medical value of THC.