Objective: To model the growth of triplets.
Method: Statistical analysis using data of a consecutive series of 76 triplet gestations (223 livebirths).
Results: The best model to describe the growth was linear regression (r = 0.76; r2 = 0.577; p < 0.0001; y = -2,858.05 + 139.66x). 34.2% of the sets had an intratriplet discordance rate of at least 25%, and 11.8% of the triplets had a maximum discordance rate over 40%.
Conclusions: (1) Triplet fetal growth is strongly related to the gestational age; (2) the growth of each triplet in a same set may considerably differ, and (3) the use of singleton curves for triplet growth assessment is inadequate.