We hypothesized that gender and intrauterine growth retardation (IUGR) have greater effects than birth order on mortality and morbidity rates of very low birth weight (< 1501 gm) twins. Neonatal data were collected on 44 pairs of twins born alive between January 1984 and December 1987. Birth weight was 1018 +/- 289 gm and gestational age was 28.1 +/- 2.5 weeks. The male/female ratio was 46:42; 24 infants had IUGR, and 64 were appropriate in size for gestational age. Of the 88 infants, 61 (69%) survived. Birth order had no effect on outcome. Female twin pairs had a longer gestation than either male twin pairs or twins with discordant sex (29.2 +/- 2.5 weeks vs 27.4 +/- 2.0 weeks and 27 +/- 3 weeks, respectively; p < 0.002). They also had a lower mortality rate (14% vs 47% and 25%; p < 0.001) and a lower incidence of bronchopulmonary dysplasia (22% vs 57% and 50%; p < 0.02). Infants with IUGR had an increased mortality rate (50% vs 23%; p < 0.02) and an increased sepsis rate (61% vs 25%; p < 0.02) compared with infants with appropriate size for gestational age who were matched for gestational age. Multiple logistic regression analysis to assess the independent effects of gestational age, gender, and IUGR on mortality rate, bronchopulmonary dysplasia, and intraventricular hemorrhage revealed that gestational age was the most significant contributor to all three outcome variables; IUGR contributed to an increased mortality rate, and male gender contributed to the occurrence of bronchopulmonary dysplasia.