SYNOPSIS
The mean error of prediction of mature height in 29 untreated tall girls was similar to that reported for normal children. There was no real tendency to overpredict or under‐predict for the group as a whole. There was, however, a real tendency to overpredict at skeletal ages of 14 years and over, and a definite tendency to underpredict in girls who were retarded skeletally. Predictions should be based on the skeletal age of the hand‐wrist, excluding the carpals, or on the skeletal age of the knee.