Random regressionmodels were used to estimate genetic parameters for test-day milk yield (PLDC)of Alpine dairy goats, implemented by Bayesian methods with Gibbs Sampling.The estimates were compared with those obtained by random regression analysis,using REML. Heritability estimates obtained by Bayesian analysis ranged from0.18 to 0.37, while those obtained by REML ranged from 0.09 to 0.32. Geneticcorrelations between yields of close test days approached the unit, but decreasedgradually as the interval between test days increased. Results indicated thatrandom regression models are appropriate to model the covariance structure ofPLDC and to predict genetic gains and select animals along the lactation trajectoryof dairy goats. Results obtained by Bayesian and REML approaches were similar,although genetic variance and heritability estimates were slightly higher withBayesian methods.
