No randomised controlled trials of treatment of twin-to-twin transfusion syndrome (TTTS) exist. Since severely preterm neonatal survival has increased over time, survival as an outcome measure is confounded by improvements in neonatal care. The diagnosis-to-delivery interval is a measure of success of treatment independent of improvements in neonatal care. We wished to evaluate whether treatment of TTTS is associated with a lengthening of the diagnosis-to-delivery interval. MEDLINE search was performed supplemented by careful reference review. All TTTS series were included where the following information on each patient was available: survival, fetal demise, gestational age at diagnosis and diagnosis-to-delivery interval in days. Inclusion criteria: gestational age at diagnosis < 29 weeks and diagnosis by ultrasound in the absence of maternal symptoms. Cases undergoing multiple types of treatment were excluded. Eight publications met inclusion criteria and included the following cases: controls (n = 16), amnioreduction (n = 61), septostomy (n = 12), and fetoscopic laser occlusion of chorioangiopagus vessels (n = 51). There was no difference in the diagnosis-to-delivery interval, overall survival, at least one survivor, or number of fetal deaths between the four groups. Logistic regression using at least one survivor as the dependent variable revealed a positive association with gestational age at diagnosis and with diagno-sis-to-delivery interval, a negative correlation with fetal death, and no correlation with treatment group. We conclude that there is no difference in diagnosis-to-delivery interval or survival for any treatment for TTTS compared to expectant management. The lack of significance appears to be due to small sample sizes.