Human growth is a nonlinear process with marked variation in growth rate during the short-term. It is not known how long-term height gain or stature is influenced by short-term changes in height and weight. This study has addressed these issues by using thrice weekly height and weight measurements during 1 year in 43 normal prepubertal children (aged 5.7-7.7 y) to construct individual height and weight velocity curves by regression analysis. The former were comprised of 3 to 6 growth spurts separated by stasis, whereas the latter were characterized by 2 to 5 periods of weight gain separated by periods of weight loss. Stepwise regression analysis to determine characteristics of these curves that influence stature and growth showed that height SD score was correlated to the mean absolute weight velocity amplitude (+), the mean length of height velocity peaks (-), and the number of periods of weight gain (-) (r2 = 38%). In contrast, change in height SD score (delta height SD score) was correlated to the number (+) and mean amplitude (+) of the periods of weight gain and the mean height velocity peak amplitude (+) (r2 = 44%). Examination of changes in height relative to weight during the year in the whole group revealed that height increased relative to weight in autumn and spring, whereas the reverse occurred during the winter months. We conclude that 1) both height and weight velocities during 1 year show a biphasic pattern, 2) there is seasonal variation in the short-term change in height relative to weight, and 3) prepubertal stature and the amount grown through the year are related to short-term changes in height and weight. Our data indicate that large but infrequent changes in weight with growth spurts of short duration are found in tall children. Good growth during the year was related to large but frequent gains in weight and large individual spurts in height.