In order to understand the effects of intrauterine growth discordancy (15% or more birth weight difference), birth weight, and gestational age on the neonatal adverse events (including 1- or 5-min Apgar scores < 7, neonatal death, ICU admission, respiratory distress, hypoglycemia, hypocalcemia, perinatal infection, blood transfusion, and hyperbilirubinemia) in third trimester twin gestations, 279 consecutive twin pairs delivered from January 1986 to December 1991 were studied. Univariate analyses showed discordant twins were smaller than concordant twins in gestational age by 1.4 weeks (35.74 and 37.14 weeks respectively). When birth weight was compared, that of smaller (one with lower birth weight in a pair) discordant twins (1,951 g) was significantly lower than that of smaller concordant twins (2,423 g), while larger (one with higher birth weight in a pair) discordant twins (2,556 g) and concordant twins (2,594 g) showed no significant difference. Univariate analysis indicated there was a tendency for adverse events to occur in discordant twins, especially in the smaller twin. Through logistic regression analysis, it was found that birth weight and gestational age, but not discordancy, are the predictors of the occurrence of adverse events. A smaller twin weighing no more than 2,000 g has a 10 times greater risk to develop an adverse event as compared to a twin with a birth weight over 2,000 g and a similar gestational age; while a twin with a gestational age of less than 34 weeks has a 5 times greater risk than one 34 weeks or over with a similar birth weight.