Using a large US nationwide database of 200 000 twin pregnancies, Lin and colleagues 1 dared to confront a difficult problem: twin pregnancies. Analyses of twin pregnancies are a challenge for perinatal epidemiologists, such as me, owing to, for example, an individual with 2 infants, chorionicity, different sexes in a pair, and discordant birth weights as well as numerous potential traps. In addition, the "physiological normal term" for twins is approximately 36 weeks, with few