The objective of the present study was to promote a better understanding of the rubber yield of young progenies of Hevea brasiliensis evaluating the stability and the repeatability of early rubber yield test employed in a genetic breeding programme. The genetic material consisted of 30 open pollination progenies evaluated in the Colina county, São Paulo state, Brazil, in a randomised block design. Rubber yield was evaluated by the Hamaker Morris-Mann early rubber yield test in 5-year-old plants and the modified Hamaker Morris-Mann test in 3- and 4-year-old plants. The mixed linear model procedure Reml/Blup (Restricted Maximum Likelihood/Best Linear Unbiased Prediction) was used in the statistical analyses. The progeny effect was significant and indicated favourable conditions for selection. The rubber yield trait was highly influenced by the environmental conditions in the different measurements. The predicted genetic gains indicated favourable conditions for selection. The repeatability coefficient (\(\hat{p}\)= 0.42) showed a moderate capacity of the genotypes to repeat the expression of the trait in the same manner from one measurement to another. Selection made on the criterion of temporal stability for early rubber yield coincided better with progenies selection in the fifth year than with selection in the third year. The predicted genetic gains indicated favourable condition for progenies selection using early yield tests at different ages. Early selection of high yield progenies in the third year, selects partially promising progenies. Therefore, when early selection is made in the third year, selection in the following years is recommended.

本研究的目的是促进对巴西橡胶树幼后代橡胶产量的更好理解。评估在遗传育种计划中使用的早期橡胶产量测试的稳定性和可重复性。遗传材料由30个开放授粉后代组成，该后代在巴西圣保罗州科利纳县进行了随机区组设计评估。橡胶产量通过Hamaker Morris-Mann在5年龄植物中的早期橡胶产量测试以及改良的Hamaker Morris-Mann在3年和4年龄植物中的测试进行评估。统计分析中使用了混合线性模型过程Reml / Blup（受限最大似然/最佳线性无偏预测）。后代效应显着，为选择提供了有利条件。在不同的测量中，橡胶的屈服性状受环境条件的影响很大。预测的遗传增益表明选择的有利条件。\（\ hat {p} \） = 0.42）表明基因型具有中等能力，可以从一种测量到另一种测量以相同的方式重复性状的表达。根据早期橡胶产量的时间稳定性标准进行的选择与第五年的后代选择相比，第三年的选择更好地符合。预测的遗传增益表明使用不同年龄的早期产量测试选择后代的有利条件。第三年尽早选择高产子代，选择部分有前途的子代。因此，在第三年进行早期选择时，建议在接下来的几年中进行选择。