The molecular mechanism of action of Δ⁹-tetrahydrocannabinol (THC), the psychotropic constituent of Cannabis, has been a puzzle during the three decades separating its characterization, in 1964, and the cloning, in the 1990s, of cannabinoid CB₁ and CB₂ receptors. However, while these latter proteins do mediate most of the pharmacological actions of THC, they do not seem to act as receptors for other plant cannabinoids (phytocannabinoids), nor are they the unique targets of the endogenous lipids that were originally identified in animals as agonists of CB₁ and CB₂ receptors, and named endocannabinoids. Over the last decade, several potential alternative receptors for phytocannabinoids, endocannabinoids, and even synthetic cannabimimetics, have been proposed, often based uniquely on pharmacological evidence obtained in vitro. In particular, the endocannabinoid anandamide, and the other most abundant Cannabis constituent, cannabidiol, seem to be the most “promiscuous” of these compounds. In this article, we review the latest data on the non-CB₁, non-CB₂ receptors suggested so far for endocannabinoids and plant or synthetic cannabinoids, and lay special emphasis on uncharacterized or orphan G-protein-coupled receptors as well as on transient receptor potential channels.