Twin-twin transfusion syndrome is a serious complication of monozygotic, monochorionic, diamniotic twins resulting from transplacental vascular communications. In this syndrome blood is thought to be shunted from one twin--donor, who develops anaemia, growth retardation and oligoamnios, to the other twin--recipient, who becomes plethoric, macrosomic and develops polyhydroamnios. The incidence of twin-twin transfusion syndrome ranges from 5-15% of all twin pregnancies. If this condition develops in the second trimester, it is usually associated with spontaneous abortion and death of one or both fetuses before viability. Developing the syndrome in the third trimester has better perinatal outcome. Mortality rates ranging from 56%-100%, depending on gestational age and severity of the syndrome. The ultrasound criterias for diagnosis, in this study, were the presence of twins of the same sex with discordant growth, with oligohydroamnios in one twin sac and polyhydroamnios in the other one, one placenta and thin membrane between twins. The present study shows clinical course of 14 cases and value of Doppler ultrasound to analyze the usefulness of umbilical artery blood flow velocimetry for predicting the risk of twin-twin transfusion syndrome. 14 twin pregnancies with twin-twin transfusion syndrome were diagnosed during the last four years period and prospectively followed. 9 cases were diagnosed before the completion od 28 weeks of gestation. The mean gestational age was 21.6 _+ 4.2 weeks at diagnosis and 23.2 +_ 3.6 weeks at delivery. 5 cases were diagnosed after 28 weeks of gestation. The mean gestational age in this group was 29.6 +_ 2.1 weeks at diagnosis and 33 +_ 3.3 weeks at delivery. The survival rate in this study was 29% (8/28). 9 cases ended in spontaneous abortion between 18th and 27th weeks of pregnancy (table 1) and 5 in premature labor (table 2). There were 7 intrauterine death (5 at admission and 2 few days after admission) and 13 neonatal deaths. Overall mortality rate was 71% (20/28). Up to 28th weeks mortality rate was 100% and after 28th weeks mortality rate was 20% (2/10). 9 cases had "stuck" twin phenomenon. The differences in the Doppler indexes from twin-twin transfusion syndrome cases significantly exceeded those without this syndrome. This difference seemed to predict the risk of twin-twin transfusion syndrome, but the number of cases is too small for general conclusions. Treatment regimens for twin-twin transfusion syndrome have included bed rest, tocolytic agents and serial amniocenthesis for decompression in some cases. Neither serial amniocenthesis nor tocolytic agents use were associated with an improved survival rate in our study. Twin-twin transfusion syndrome can be diagnosed early in second trimester of pregnancy, but the survival rate remains low with the current methods of treatment. Survival was related to gestational age at delivery and birth weight. Twin-twin transfusion syndrome diagnosed before 28th weeks of gestation represents one of the most lethal conditions in perinatal medicine today.