Obsessive-compulsive disorder (OCD) is characterised by structural and functional deficits in the cortico-striato-thalamic-cortical (CSTC) circuitry and abnormal neurochemical changes are thought to modulate these deficits. The hypothesis that an imbalanced concentration of the brain neurotransmitters, in particular glutamate (Glu) and gamma-amino-butyric acid (GABA), could impair the normal functioning of the CSTC, thus leading to OCD symptoms, has been tested in humans using magnetic resonance spectroscopy (MRS) and positron emission tomography (PET). This chapter summarises these neurochemical findings and represents an attempt to condense such scattered literature. We also discuss potential challenges in the field that may explain the inconsistent findings and suggest ways to overcome them. There is some convergent research from MRS pointing towards abnormalities in the brain concentration of neurometabolite markers of neuronal integrity, such as N-acetylaspartate (NAA) and choline (Cho). Lower NAA levels have been found in dorsal and rostral ACC of OCD patients (as compared to healthy volunteers), which increase after CBT and SSRI treatment, and higher Cho concentration has been reported in the thalamus of the OCD brain. However, findings for other neurometabolites are very inconsistent. Studies have reported abnormalities in the concentrations of creatine (Cr), GABA, glutamate (Glu), glutamine (Gln), Ins (myo-inositol), and serotonin (5-HT), but most of the results were not replicated. The question remains whether the NAA and Cho findings are genuinely the only neurochemical abnormalities in OCD or whether the lack of consistent findings for the other neurometabolites is caused by the lower magnetic field (1-3 Tesla (T)) used by the studies conducted so far, their small sample sizes or a lack of proper control for medication effects. To answer these questions and to further inform the biological underpinning of the symptoms and the cognitive problems at the basis of OCD we need better controlled studies using clear medicated vs unmedicated groups, larger sample sizes, stronger magnetic fields (e.g. at 7 T), and more consistency in the definition of the regions of interest.