There is consistent evidence that psychosis is associated with a degree of peripheral immune activation. Many studies and meta-analyses report increased circulating concentrations of pro-inflammatory cytokines including IL-6, IL-1β and TNF-α together with acute phase proteins, such as CRP in patients with psychosis compared with controls. Meta-analysis confirms increased circulating IL-6 and other inflammatory markers in medication-naïve first-episode psychosis (FEP) (1), and in the CSF of schizophrenia patients compared with controls (2). Longitudinal studies show an association between elevated IL-6/CRP in childhood/adolescence and risk of psychotic symptoms or diagnosis of schizophrenia in adulthood (3,4). Genetic analysis shows that possession of a functional variant in the IL-6 R gene (Asp358Ala, rs2228145), which is known to reduce the activity of IL-6, is also associated with decreased risk of psychosis (5). This suggests that the association of raised IL-6 with psychosis may be causal and not due to reverse causality or residual confounding, i.e., the effect of a factor associated with psychosis that also by chance increases levels of this cytokine. However, Mendelian randomization analysis also suggests that elevated CRP levels may be protective for schizophrenia (6), in contrast to observational studies consistently reporting higher CRP levels in patients with the illness compared with controls (7). Divergent results for two known pro-inflammatory markers raises questions about potential mechanisms though which immune dysfunction may influence brain and behaviour to increase the risk of psychotic disorders. One explanation is that genetically predicted levels of low CRP may predispose to infections, which in turn, may increase schizophrenia risk through immune and non-immune mechanisms (6,8).