Cannabis has been used to treat pain for thousands of years. However, since the early part of the 20th century, laws restricting cannabis use have limited its evaluation using modern scientific criteria. Over the last decade, the situation has started to change because of the increased availability of cannabis in the United States for either medical or recreational purposes, making it important to provide the public with accurate information as to the effectiveness of the drug for joint pain among other indications. The major psychotropic component of cannabis is Δ9-tetrahydrocannabinol (THC), one of some 120 naturally occurring phytocannabinoids. Cannabidiol (CBD) is another molecule found in herbal cannabis in large amounts. Although CBD does not produce psychotropic effects, it has been shown to produce a variety of pharmacological effects. Hence, the overall effects of herbal cannabis represent the collective activity of THC, CBD and a number of minor components. The action of THC is mediated by two major G-protein coupled receptors, cannabinoid receptor type 1 (CB1) and CB2, and recent work has suggested that other targets may also exist. Arachidonic acid derived endocannabinoids are the normal physiological activators of the two cannabinoid receptors. Natural phytocannabinoids and synthetic derivatives have produced clear activity in a variety of models of joint pain in animals. These effects are the result of both inhibition of pain pathway signalling (mostly CB1) and anti-inflammatory effects (mostly CB2). There are also numerous anecdotal reports of the effectiveness of smoking cannabis for joint pain. Indeed, it is the largest medical request for the use of the drug. However, these reports generally do not extend to regulated clinical trials for rheumatic diseases. Nevertheless, the preclinical and human data that do exist indicate that the use of cannabis should be taken seriously as a potential treatment of joint pain.