To investigate the association between maternal imprisonment during pregnancy and perinatal outcomes. Systematic review and meta-analysis. Prisons in developed countries. Imprisoned pregnant women. Two reviewers extracted the data independently according to a fixed protocol. Studies were included if they were cohort or case-control studies with women identified as being imprisoned at any point during pregnancy and if they included a comparison group of women who had not been imprisoned. Case series without a comparison group of women who were not imprisoned were excluded, as were studies that did not include information on the pre-specified outcomes. Perinatal or infant death, stillbirth, preterm delivery, low birthweight, fetal anomalies and use of breast milk substitutes. Of 28 relevant papers, 10 met the inclusion criteria, involving 1960 imprisoned pregnant women and 10,858 controls. There were significant differences in results between studies comparing imprisoned women with population controls and those making comparisons with disadvantaged control women. Imprisoned women are more likely to deliver prematurely and have a low birthweight baby than population control women. However, when compared with a similarly disadvantaged group, imprisoned woman are less likely to have a stillbirth or low birthweight baby, suggesting imprisonment may have a beneficial effect. Particular perinatal outcomes appear to be improved in imprisoned women compared with similarly disadvantaged women. Imprisonment of the mother has a beneficial effect on the birthweight of her baby.