In recent times, several longitudinal studies aimed at clarifying whether cannabis use during adolescence might play a causal role in the subsequent risk of developing bipolar disorder have been published. Although their methodological heterogeneity precludes any meta-analytic approaches, evidence from these studies can be systematically evaluated using the Bradford Hill criteria. A biological gradient is supported by evidence on the dose–response relationship between exposure severity and outcome. As such, the effect of cannabis use on bipolar disorder onset is likely to be strong, coherent, plausible, and based on a clear temporality. In addition, some analogies can be hypothesized between studies testing the possible causal role of cannabis in the development of bipolar disorder and those is schizophrenia. Cannabis may represent a precipitating agent inducing bipolar disorder in a multicausal model of individual vulnerability. However, this relationship seems to be only partially consistent and nonspecific, and the experimental evidence is strongly suggestive but, as yet, inconclusive. Nevertheless, in summary, it seems there is sufficient support for the hypothesis that cannabis use during adolescence may play a causal role in bipolar disorder, although further studies are needed to consolidate the evidence.