A total of 36212 racing performance data recorded between 1990 and 2006 from 2325 Trotter horses in Spain were studied by Random Regression Model (RRM), in order to estimate the genetic parameters for race performance in young horses (2-4 years old) of this breed by distances ranging between 1600-2750m. The pedigree of each horse was extended until the fourth generation, with a total of 9201 individuals. Racing time per kilometre over the trajectory of distances was the dependent variable. The hippodrome-date of race (405), sex (3: male, female and gelding) and a second order Legendre polynomial were the fixed effects. The animals (9201), the driver (1007) and the permanent environmental effects due to repetitions of records from the same animals were incorporated as random with a Legendre polynomial of order 1 in the model. According to our results, the additive variance components and the heritability (h2) decrease as race distance increase, (ranging between h2=0.32 at 1600 m and h2=0.12 at 2700 m). The genetic correlations between close distances were higher. With these results, it was possible to estimate the Breeding Value (BV) for all animals and all distances represented in the data set. A very high variability was estimated for the BV and important differences were detected in the form of responses of the horses along the trajectory of distances. The use of RRM is highly recommended in the breeding programs of racing performance of the Trotter horse in Spain.