<p>Male strobili production in some conifers is unstable, fluctuating from year to year. For breeding and seed production, evaluation of clonal variation in male strobili production appropriately, considering yearly fluctuations, is necessary. In this study, the number of branches and number of years required for evaluation were studied by investigating the male strobili setting via gibberellin treatment of first-generation plus tree clones for six years in <i>Chamaecyparis obtusa</i>. The analysis of various variance components showed that among-clone variance was the largest, followed by among-year variance, while the variance of ramets and branches was estimated to be small. The simulation results showed that surveying three branches has a correlation coefficient of 0.945 with surveying six branches, suggesting that surveying three to four branches per clone is sufficient. Simulation results show that the results of the five-year study have a correlation coefficient of 0.96 with the results of the six-year study, and that the clone ranking is also stable. Notably, clonal values were relatively lower at younger ages (&lt; 10 years) than at older ages (&gt; 25 years), indicating the possibility of a slight underestimation at younger ages. This study may provide useful guidelines for investigating male fecundity in conifers with high annual variability.</p>