Psychosis is a mental health syndrome presenting with hallucinations - sensory perception of stimuli not objectively present in the environment, delusions - false beliefs held with firm conviction, and disorganised thinking and speech. These experiences occur on a spectrum of severity. People at Clinical High Risk for Psychosis (CHRp) have sub-threshold psychotic symptoms (for example, delusions which the person is able to question). Around 25% of individuals at CHRp develop frank psychosis in the three years after their initial presentation to mental health services. This is markedly higher than the approximate 1% prevalence of psychosis in the general population. The symptoms people at CHRp experience can be distressing and may affect people's engagement in professional and social activities. Thus, an increasing amount of research into psychosis development and its symptomatic spectrum has focused on this group. Despite extensive research, the biological basis of psychosis remains largely unknown. A promising avenue of study is the investigation of emotion processing along a psychosis spectrum. Difficulties with emotion expression and less accurate identification of emotion- related stimuli were reported across the psychosis spectrum. These phenomena were shown to be associated with people's mental health and life outcomes. Therefore, there is an urgent need to understand the neural mechanisms underlying these processes to develop new treatments and prevention strategies for psychosis. Functional magnetic resonance imaging (fMRI) allows to visualise regional brain activation as an individual performs a task containing emotion-related stimuli. Such studies in chronic psychotic disorders consistently showed different brain activation to emotion-related stimuli compared to healthy controls. It is less well understood whether the same pattern can be found in individuals experiencing their first episode of psychosis (FEP) and in people at CHRp. Moreover, the molecular mechanisms which may underlie such abnormalities are still under investigation. The identification of these processes may inform the development of new pharmacological treatments for psychosis. Gamma-aminobutyric acid (GABA) is the main inhibitory neurotransmitter in the brain which orchestrates neural network activity. As difficulties with emotion processing seen in psychosis are associated with altered recruitment of certain brain regions, examining GABAergic neurotransmission may help understand the underlying molecular mechanisms. Positron emission tomography (PET) is the only non- invasive method of studying regional availability of GABAergic receptors in the brain in vivo. Thus, it is a valuable method of investigating GABAergic function in the psychosis spectrum. This thesis combines the study of emotion processing with fMRI, and of GABAergic neurotransmission with PET, in individuals at CHRp and in people with an FEP. It includes a systematic review and meta-analysis of fMRI studies of emotion processing in these populations. It also describes an experiment aiming to determine the most accurate way of modelling brain activation to emotion-related stimuli with fMRI. Further, it recounts a project exploring the contribution of distinct components of the GABAergic system (individual subunits of the GABA receptor type A and molecular markers of specific GABAergic neuron subtypes) to the signal obtained with GABA PET. Finally, it presents preliminary findings from a new multimodal PET-fMRI study investigating the relationship between emotion processing and GABAergic signalling in individuals at CHRp and in people with an FEP. The thesis closes with a discussion bringing together the results of these studies and putting them in context of the current conceptualisations of psychosis pathophysiology.