We evaluated 56 twin pregnancies representing the tenth decile of the mean twin birth weight distribution to investigate whether larger twins face the same increased perinatal risk as do macrosomic singletons. Compared with pregnancies in the ninth decile, no significant difference was found between the means of maternal age, parity, and gestational age; between the rates of presentation combinations and cesareans; or between the neonatal sex ratios. However, the incidence of growth-discordant pairs was significantly higher in the heavier group (P = .034). Compared with the general twin population, the tenth-decile group contained significantly fewer primiparas (P = .0023). Maternal obesity and diabetes were infrequent and could not explain the increased birth weight. The neonatal outcome was excellent. Although the comparison revealed insignificant differences, it is possible that the combination of higher parity, malpresentation rate, and male-to-female ratio may be operative in the genesis of large twins.