Cannabis is the most widely used illicit drug. Despite this, only a small number of studies have investigated the long-term neurotoxic consequences of cannabis use. Structural and functional neuroimaging techniques are powerful research tools to investigate possible cannabis-induced pathophysiological changes. A computer literature review was conducted in the MEDLINE and PsycLIT databases between 1966 and November of 2004 with the search terms 'cannabis', 'marijuana', 'neuroimaging', 'magnetic resonance', 'computed tomography', 'positron emission tomography', 'single photon emission computed tomography", 'SPET', 'MRI' and 'CT'. Structural neuroimaging studies have yielded conflicting results. Most studies report no evidence of cerebral atrophy or regional changes in tissue volumes, and one study suggested that long-term users who started regular use on early adolescence have cerebral atrophy as well as reduction in gray matter. However, several methodological shortcomings limit the interpretation of these results. Functional neuroimaging studies have reported increases in neural activity in regions that may be related with cannabis intoxication or mood-change effects (orbital and mesial frontal lobes, insula, and anterior cingulate) and decreases in activity of regions related with cognitive functions impaired during acute intoxication. The important question whether residual neurotoxic effects occur after prolonged and regular use of cannabis remains unclear, with no study addressing this question directly. Better designed neuroimaging studies, combined with cognitive evaluation, may be elucidative on this issue.