The temporal dynamics and extent of additive genetic control of the ability to produce secondary leader growth (SLG) were examined in a Pinus pinaster (Ait.) progeny trial and the benefits of including this binary ''trait'' (presence/absence) in a selection scheme for improved height growth were assessed. Trees with SLG had superior height growth, especially at younger ages (< 11 years). An average of 39% of the trees had SLG, but differences between years and families were manifest. The ability to form SLG was under fairly strong additive genetic control (h2 = 0.38) compared with the annual height increment (AHI) (h2 = 0. 17). Trees with SLG in one year were more likely to produce SLG in later years. However, the consistency of producing SLG deteriorated rapidly with time and became nonsignificant after 4-6 years. In young trees ( < 10 years of age), the frequency of SLG and mean AHI were positively correlated at the family level. In older trees (> 14 years of age), the same correlations were mostly negative. Including occurrence of SLG as an auxiliary selection trait (with zero economic value) in an index with total annual height growth improved overall gain expectations for age 2 to age 22 height growth. However, according to a stochastic risk analysis, the benefit of using occurrence of SLG as an auxiliary trait was more than offset by an unacceptable increase in the standard error of the gain predictions as compared to the gain and associated errors of predicting gain from height growth alone. Stochastically, 10 years of age appears to be the most efficient age for family selections for height growth.