Until recently, it was possible to state with some confidence that there was no evidence of cannabis-related brain damage in humans. There was some support from the animal literature, but few human studies had been conducted where the findings could not be explained by methodological or other confounding factors. Recent evidence for gross morphological, connectivity and microstructural changes has now emerged that warrants further consideration. If cannabis were found to alter the structural integrity of the brain, then this may assist us to understand the mechanisms by which cannabis triggers psychotic symptoms or overt psychosis in vulnerable individuals. Evidence from animal studies Cannabinoids, either endogenous or exogenous, possess both neuroprotective and neurotoxic properties (Sarne and Mechoulam, 2005; Kano et al., 2009). Cannabinoid-receptor activation induces morphological changes to neurons, such as inhibition of new synapse formation (Kano et al., 2009), and at crucial neurodevelopmental stages (prenatal and adolescent), exposure to cannabinoids impacts on neural cell survival and maturation (Chapters 6, 7) (Downer and Campbell, 2010). The role of different cannabinoids in controlling neural-cell survival or death is a complex issue that is influenced by the dose, duration of exposure and route of administration, but also the neural-cell type and its stage of differentiation (Downer and Campbell, 2010). Contradictory hypotheses circulate regarding the doses of Δ9-tetrahydrocannabinol (THC) that may be neurotoxic or neuroprotective. Some suggest that single high doses of THC are neuroprotective within a limited timeframe, but that low doses are neurotoxic and, with chronic exposure, induce neuronal death (Sarne and Keren, 2004; Tselnicker et al, 2007; Sarne and Mechoulam, 2005). However, large doses of THC applied directly to cultured hippocampal neurons, and both high and low doses to cultured cortical neurons, have been shown to cause cell death or significant neurotoxic changes (eg. shrinkage of cell bodies and DNA-strand breaks) characteristic of neuronal apoptosis (Chan et al., 1998; Campbell, 2001; Downer et al., 2001). Indeed, even a single administration of an ultra-low dose of THC (0.001–0.002 mg/kg) has been shown to result in long-term cognitive impairment (in spatial learning, strategy and working memory) in mice. These defecits persisted for at least 5 months post-injection and were associated with activation of extracellular-regulated kinase (ERK) in the cerebellum and hippocampus (Tselnicker et al., 2007; Amal et al., 2010). The authors suggested that low THC concentration is the main determinant of long-lasting neuronal effects following chronic exposure to cannabinoids, due to their slow clearance and accumulation (Amal et al., 2010).