High levels of psychiatric morbidity in prisoners have been documented in many countries, but it is not known whether rates of mental illness have been increasing over time or whether the prevalence differs between low-middle-income countries compared with high-income ones. To systematically review prevalence studies for psychotic illness and major depression in prisoners, provide summary estimates and investigate sources of heterogeneity between studies using meta-regression. Studies from 1966 to 2010 were identified using ten bibliographic indexes and reference lists. Inclusion criteria were unselected prison samples and that clinical examination or semi-structured instruments were used to make DSM or ICD diagnoses of the relevant disorders. We identified 109 samples including 33 588 prisoners in 24 countries. Data were meta-analysed using random-effects models, and we found a pooled prevalence of psychosis of 3.6% (95% CI 3.1-4.2) in male prisoners and 3.9% (95% CI 2.7-5.0) in female prisoners. There were high levels of heterogeneity, some of which was explained by studies in low-middle-income countries reporting higher prevalences of psychosis (5.5%, 95% CI 4.2-6.8; P = 0.035 on meta-regression). The pooled prevalence of major depression was 10.2% (95% CI 8.8-11.7) in male prisoners and 14.1% (95% CI 10.2-18.1) in female prisoners. The prevalence of these disorders did not appear to be increasing over time, apart from depression in the USA (P = 0.008). High levels of psychiatric morbidity are consistently reported in prisoners from many countries over four decades. Further research is needed to confirm whether higher rates of mental illness are found in low- and middle-income nations, and examine trends over time within nations with large prison populations.