ABSTRACT:The aim of this study was to verify the importance of genotype-environment interaction (GEI) for age at first calving (AFC), scrotal circumference (SC) and yearling weight (YW) in Nellore cattle, using reaction norms in multi-trait random regression models and analyze the efficiency of selection for AFC based on SC and YW in different environments. The research was conducted using 28,871, 41,386 and 89,152 records for AFC, SC and YW, respectively. Environmental groups (EGs) were defined based on farm and year of birth and management group (at birth, weaning and yearling) for all traits. For YW sex of animal was added to EGs. Average post-weaning weight gain was used to assess environmental conditions. Models for SC and YW included contemporary group and the covariate, age of animal at evaluation, as fixed effects, and the residual and direct additive genetic effects as random effects. The same model was used for AFC, excluding age of animal effects. The (co)variance components were estimated using Wombat software. The additive genetic and phenotypic variances estimates for AFC, SC and YW increased as the environmental conditions improved, with a greater intensity for AFC. The heritability estimates for all traits increased as the environment became more favorable and ranged from low to medium for AFC (0.04 to 0.48), high for SC (0.51 to 0.67) and medium to high for YW (0.23 to 0.76). The genetic correlation estimates between AFC and SC, AFC and YW, YW and SC in different environments, ranged from -0.15 to -0.68, -0.18 to -0.35, -0.1 to +0.68, respectively. The efficiency of indirect selection for AFC using YW and SC ranged from 4 to 395 and from 12 to 333, respectively. There is an important GEI effect for AFC, SC and YW traits, so the expected genetic gain for these traits depend on the environment in which the animals are exploited. The SC and YW traits can be used as a selection criterion for sexual precocity in unfavorable environments.